calibration of line scales final report - bipm€¦ · roman fira +421 2 602 94 321 +421 2 654 29...
Post on 13-Jul-2020
0 Views
Preview:
TRANSCRIPT
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 1
EUROMET.L-K7-2006_Final
Metrology Institute of the Republic of Slovenia
University of Maribor Faculty of Mechanical Engineering
Laboratory for Production Measurement
EUROMET Key Comparison, EUROMET.L-K7
(EUROMET Project 882)
Calibration of line scales
Final Report
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 2
EUROMET.L-K7-2006_Final
Contents
1 INTRODUCTION .................................................................................................. 3
2 ORGANISATION .................................................................................................. 4
2.1 Participants ................................................................................................... 4
2.2 Linking laboratories ...................................................................................... 8
2.3 Form of comparison ...................................................................................... 8
2.4 Circulation of the artefact and performance of the measurements ............... 8
2.5 Transport of the artefacts ........................................................................... 10
3 THE ARTEFACT ................................................................................................. 11
3.1 Description of the artefact ........................................................................... 11
3.2 Fixing the artefact ....................................................................................... 12
4 MEASUREMENT INSTRUCTIONS .................................................................... 13
4.1 Traceability ................................................................................................. 13
4.2 Measurand.................................................................................................. 13
4.3 Measurement instructions .......................................................................... 14
4.4 Measurement uncertainty ........................................................................... 14
5 MEASUREMENT EQUIPMENT AND METHODS USED BY THE PARTICIPANTS .......................................................................................... 17
6 STABILITY OF THE STANDARDS ..................................................................... 20
6.1 Stability of the total line scale length ........................................................... 20
6.2 Stability of single line centres ..................................................................... 21
6.3 Impact of line scale instability ..................................................................... 24
7 ANALYSIS OF THE RESULTS ........................................................................... 25
8 MEASUREMENT RESULTS ............................................................................... 26
8.1 Measurement results- Group 1 ................................................................... 26
8.2 Measurement results – Group 2 ................................................................. 28
8.3 Correction of the original results ................................................................. 32
8.4 Link between the groups ............................................................................ 33
8.5 Intercomparison results for linked groups considering line scale stability ... 34
9 CONCLUSIONS .................................................................................................. 53
10 ACKNOWLEDGMENT ........................................................................................ 53
11 REFERENCES ................................................................................................... 53
APPENDIX 1: INTERCOMPARISON RESULTS FOR THE CASE OF SEPARATE TREATMENT OF GROUPS 1 AND 2 (WITHOUT LINK) .................. 54
APPENDIX 2: DESCRIPTIONS OF MEASUREMENT SETUPS AND METHODS AS REPORTED BY THE PARTICIPANTS .................... 75
APPENDIX 3: REPORTED UNCERTAINTY BUDGETS ....................................... 128
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 3
EUROMET.L-K7-2006_Final
1 Introduction
1.1 At its meeting in October 2005, the TC for Length identified several EUROMET key comparisons in the field of dimensional metrology. In particular, it decided that a key comparison on line standards shall be carried out. This comparison follows the Nano3 comparison (WGDM-7 preliminary comparison on nanometrology).
1.2 Due to the large number of the participants, it has been decided to have 2 groups in the project. The participants for the 2 groups were chosen in accordance with their geographical position (in order to minimize travel times and expenses for the transportation of the standards). Linking laboratories between the groups were chosen among participants in Nano3 project.
1.3 The standards for the comparison were defined at the TCL meeting in October 2005. It was decided that only one line scale of 100 mm with line distance of 0,1 mm would be measured. The 2 groups have got equal standards offered (and produced) by NPL.
1.4 The pilot laboratory for both loops of the comparison was Metrology Istitute of the Republic of Slovenia (MIRS/UM-FS). In addition, METAS (CH) and NPL (UK) were appointed as linking laboratories between the 2 groups.
1.5 A goal of the EUROMET key comparisons for topics in dimensional metrology is to demonstrate the equivalence of routine calibration services offered by NMIs to clients, as listed in Appendix C of the Mutual Recognition Agreement (MRA) [BIPM, 1999]. Therefore, participants in this comparison agreed to use the same apparatus and methods as routinely applied to client artefacts.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 4
EUROMET.L-K7-2006_Final
2 Organisation
2.1 Participants
The project began with 31 participants – 22 from EURAMET, 4 from SIM (3 NORAMET, 1 SURAMET), 2 from COOMET and 3 from APMP. Due to the large number of participants they were divided into two groups (Table 1 and Table 2). During the project, Norwegian Metrology Service (JV) and FPS Economy - DG Quality and Safety Metrology Division (SMD) from Belgium cancel their participation due to problems with equipment. In year 2006 Thailand sent a request to participate and the request was approved at the TCL meeting in October 2006. Thailand was placed in group 2.
Table 1: Participants in the group 1
Laboratory Address Contact person/tel/fax/e-mail
BEV
Bundesamt für Eich – und Vermessungswesen Arltgasse 35 AT-1160 Wien Austria
Michael Matus +43 1 49 110 540 +43 1 49 20 875 michael.matus@bev.gv.at
DZM-FSB
University of Zagreb Faculty of Mechanical Eng. and Naval Architecture Ivana Lucica 5 HR-10000 Zagreb Croatia
Vedran Mudronja +385 1 616 83 35 +385 1 616 85 99 vedran.mudronja@fsb.hr
GUM
Central Office of Measures ul. Elektoralna 2 PL-00950 Warszawa Poland
Zbigniew Ramotowski +48 22 581 9543 +48 22 620 8378 length@gum.gov.pl
INM
National Institute of Metrology Sos. Vitan-Barzesti 11 Sector 4 Bucharesti 042122-RO Romania
Alexandru Duta +40 21 334 55 20 +40 21 334 55 33 alexandru.duta@inm.ro
JV
Norwegian Metrology Service Fetvejen 99 NO-2007 Kjeller Norway
Helge Karlsson +47 64 84 84 84 +47 64 84 84 85 helge.karlsson@justervesenet.no
LNMC
Latvian National Metrology Centre 157, K. Valdemara Str. LV-1013 Riga Latvia
Edite Turka +371 7 362 086 +371 7 362 805 edite.turka@lnmc.lv
METAS
Bundesamt für Metrologie Lindenweg 50 CH-3084 Wabern Switzerland
Felix Meli +41 31 32 33 346 +41 31 32 33 210 felix.meli@metas.ch
MIKES
Centre for Metrology and Accreditation Tekniikantie 1 P.O. Box 9 FI-02151 Espoo Finland
Antti Lassila +358 10 6054 413 +358 10 6054 499 antti.lassila@mikes.fi
MIRS
University of Maribor Faculty of Mechancal Engineering Smetanova 17 SI-2000 Maribor Slovenia
Bojan Acko +386 2 220 7581 +386 2 220 7990 bojan.acko@uni-mb.si
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 5
EUROMET.L-K7-2006_Final
Laboratory Address Contact person/tel/fax/e-mail
NCM
National Centre of Metrology 52B G.M. Dimitrov Blvd. BG-1040 Sofia Bulgaria
Veselin Gavalyugov +359 2 97 02 760 +359 2 97 02 719 v.gavalyugov@bim.government.bg
NML
National Metrology Laboratory Enterprice Ireland Campus Glasnevin IE-Dublin 9 Ireland
Howard McQuoid +353 1 808 2657 +353 1 808 2026 howard.mcquoid@enterprise-ireland.com
NPL
National Physical Laboratory Hampton Road Teddington, Middlesex TW 11 OLW United Kingdom
Michael McCarthy +44 20 8943 6655 +44 20 8614 0453 michael.mccarthy@npl.co.uk
NSCIM
National Scientific Center "Institute of metrology" Myronosytskaja st., 42, Kharkov, 61002, Ukraine
Valentin Solovyov +380 57 704-98-77 +380 57 700-34-47 solovyov@metrology.kharkov.ua
OMH
National Office of Measures Németvölgyi út 37-39 H-1124 Budapest XII. Hungary
Edit Banreti +36 1 458 59 97 +36 1 458 59 27 e.banreti@omh.hu
PTB
Physikalisch-Technische Bundesanstalt Department 5.2, Length and Angle Metrology Bundesallee 100 DE-38116 Braunschweig Germany
Harald Bosse +49 531 5925200 +49 531 5925205 harald.bosse@ptb.de
SMU
Slovak Institute of Metrology Karloveská 63 SK-842 55 Bratislava Slovakia
Roman Fira +421 2 602 94 321 +421 2 654 29 592 fira@smu.gov.sk
ZMDM
Bureau of Measures and Precious Metals Mike Alasa 14 YU - 11 000 Beograd Serbia
Slobodan Zelenika +381 11 20 24 418 +381 11 21 81 668 zelenika@szmdm.sv.gov.yu
Changes in group 1 during the project:
JV
Norwegian Metrology Service Fetvejen 99 NO-2007 Kjeller Norway
Participation cancelled
CMI
Czech Metrology Institute V Botanice 4 CZ 150 72 Praha 5 Czech Republic
After measuring in Group 2 (October 06) they improved measurement capabilities. New measurements in Group 1 were approved by Euromet TCL. Results in group 2 were cancelled.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 6
EUROMET.L-K7-2006_Final
Table 2: Participants in the group 2
Laboratory Address Contact person/tel/fax/e-mail
CEM
Centro Espanol de Metrologia Alfar, 2 ES-28760 Tres Cantos (Madrid) Spain
Emilio Prieto +34 91 807 47 16 +34 91 807 48 07/809 eprieto@cem.mityc.es
CENAM
CENAM-Centro Nacional de Metrologia Division de Metrologia Dimensional Km 4,5 Carretera a Los Cues, El Marqués 76241 Queretaro Mexico
Carlos Colin Miguel Viliesid Alonso +52 442 211 05 74 +52 442 211 05 77 ccolin@cenam.mx mviliesi@cenam.mx
CMI
Czech Metrology Institute V Botanice 4 CZ 150 72 Praha 5 Czech Republic
Petr Balling +420 257 288 326 +420 257 328 077 pballing@cmi.cz
EIM
Hellenic Institute of Metrology Industrial Area of Thessaloniki Block 45 GR-57 022 Sindos Thessaloniki Greece
Christos Bandis +30 2310 56 99 99 +30 2310 56 99 96 bandis@eim.org.gr
INMETRO
Instituto Nacional de Metrologia, Normalização e Qualidade IndustrialLaboratório de Metrologia Dimensional - Lamin - Prédio 3Av. Nossa Senhora das Graças, 50 Xerém - 25250-020 Duque de Caxias Rio de Janeiro, Brazil
João Antônio Pires Alves +55 21 2679 9107 +55 21 2679 1505 jaalves@inmetro.gov.br
INRIM
Instituto Nazionale di Ricerca Metrologica (INRIM) Strada delle Cacce, 73 IT-10135 Torino Italy
Gian Bartolo Picotto +39 011 3977 469/473 +39 011 3977 459 g.picotto@inrim.it
METAS
Bundesamt für Metrologie Lindenweg 50 CH-3084 Wabern Switzerland
Felix Meli +41 31 32 33 346 +41 31 32 33 210 felix.meli@metas.ch
NIM
National Institute of Metrology Length Division Beisanhuandonglu 18 100013 Beijing China
Sitian Gao Tel: +86 10 84251574 Fax: +86 10 64218703 gaost@nim.ac.cn
NIST
National Institute of Standards and Technology Manufacturing Engineering Laboratory Precision Engineering Division Nano-Scale Metrology Group 100 Bureau Drive, Stop 8212 Bldg. 220, Rm A117 Gaithersburg, Maryland 20899-8212 USA
William B. Penzes +301 975 3477 +301 869 0822 william.penzes@nist.gov
NMi-VSL BV
NMi Van Swinden Laboratorium B.V. Thijsseweg 11 P.O. Box 654 NL-2600 AR Delft The Netherlands
Gerard Kotte +31 15 269 16 01 +31 15 261 29 71 gkotte@nmi.nl
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 7
EUROMET.L-K7-2006_Final
Laboratory Address Contact person/tel/fax/e-mail
NPL
National Physical Laboratory Hampton Road Teddington, Middlesex TW 11 OLW United Kingdom
Michael McCarthy +44 20 8943 6655 +44 20 8614 0453 michael.mccarthy@npl.co.uk
NPLI
National Physical Laboratory Physico-Mechanical Standards Length & Dimension Standards New Delhi -110012 India
R.P. Singhal +91-11-25732965 +91-11-25732965 singhal@mail.nplindia.ernet.in
NRC
Institute for National Measurement Standards (INMS) National Research Council Canada (NRC) 1200 Montreal Road Ottawa, ON, Canada K1A OR6
Jim Pekelsky +613 993 7578 +613 952 1394 jim.pekelsky@nrc.ca
SMD
FPS Economy DG Quality and Safety Metrology Division (SMD) Boulevard du Roi Albert II, 16 BE 1000 Brussels Belgium
Hugo Piree +32 2 277 7610 +32 2 277 5405 hugo.piree@mineco.fgov.be
A*Star - NMC
National Metrology Centre A*Star 1 Science Park Drive Singapore 118221
Siew Leng Tan +65 6279 1938 +65 6279 1994 tan_siew_leng@nmc.a-star.edu.sg
VNIIM
VNIIM - All-Russian Institute for Metrology19 Moscovsky prosp.RU - 198005 St. PetersburgRussia
Konstantin V.Chekirda +7 812 323 9664 +7 812 713 0114 K.V.Chekirda@vniim.ru
Changes in group 2 during the project:
SMD
FPS Economy DG Quality and Safety Metrology Division (SMD) Boulevard du Roi Albert II, 16 BE 1000 Brussels Belgium
Participation cancelled
CMI
Czech Metrology Institute V Botanice 4 CZ 150 72 Praha 5 Czech Republic
Results in group 2 were cancelled New measurements in group 1
NIMT
National Institute of Metrology Thailand Department of Dimensional Metrology 3/5 Moo 3, Klong 5, Klongluang, Pathumthani 12120 Thailand
NEW PARTICIPANT (approved at Euromet TCL meeting 2006)
Contact: Anusorn Tonmueanwai +662 577 5100 ext 1216 +662 577 3658 / 662 5773659 anusorn@nimt.or.th
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 8
EUROMET.L-K7-2006_Final
2.2 Linking laboratories
Linking laboratories between the two groups were METAS – CH and NPL – UK. The linking laboratories measured both artefacts in the beginning and at the end of the loop. Their measurements were also used for evaluating the stability of the artefacts.
2.3 Form of comparison
2.3.1 The comparison was performed in a ‘circular’ form in both groups. The artefact was circulated within a group of laboratories. Before sending it to the non-EU participants (in the end of each loop), they were returned to the pilot laboratory in order to prepare the necessary ATA Carnet and other forms for the custom formalities.
2.3.2 All results were communicated directly to the pilot laboratory.
2.4 Circulation of the artefact and performance of the measurements
2.4.1 The participating laboratories were asked to specify a preferred timetable slot for their measurements of the artefact - the timetables given below have been drawn up taking these preferences into account.
Table 3: Time schedule for the group 1
Laboratory Country Date
MIRS Slovenia July 2006
METAS Switzerland August 2006
NPL United kingdom September 2006
OMH Hungary October 2006
BEV Austria November 2006
SMU Slovakia December 2006
PTB Germany January 2007
GUM Poland February 2007
MIKES Finland March 2007
JV Norway April 2007
LNMC Latvia May 2007
NML Ireland June 2007
NCM Bulgaria July 2007
INM Romania August 2007
ZMDM Serbia September 2007
DZM-FSB Croatia October 2007
NSCIM Ukraine November 2007
METAS Switzerland December 2007
NPL United kingdom January 2008
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 9
EUROMET.L-K7-2006_Final
Table 4: Time schedule for the group 2
Laboratory Country Date
METAS Switzerland August 2006
NPL United kingdom September 2006
CMI Czech Republic October 2006
EIM Greece November 2006
INRIM Italy December 2006
NMi-VSL Netherlands January 2007
CEM Spain February 2007
SMD Belgium March 2007
INMETRO Brazil April 2007
CENAM Mexico May 2007
NIST USA June 2007
NRC Canada July 2007
A*Star–NMC Singapore August 2007
NIM China September 2007
NPLI India October 2007
NPL United kingdom November 2007
METAS Switzerland December 2007
VNIIM Russia January 2008
2.4.2 Each laboratory had one month for calibration and transportation. With its confirmation to participate, each laboratory has confirmed that it was capable to perform the measurements in the time allocated to it. In this way it was assured, that the artefact arrived in the country of the next participant at the beginning of the next month. However, due to many customs problems (ATA not stamped, ATA lost, substitutional ATA, special arrangements – temporary imports for countries without ATA agreement, …), cancellation of participation, new laboratory, additional measurement in another group, some measurements were not performed in the original time schedule. Changes are indicated in Table 5 and 6.
Table 5: Changes in measurement schedule in group 1
Laboratory Country Date
ZMDM Serbia October 2007
DZM-FSB Croatia December 2007
NSCIM Ukraine January 2007
CMI Czech Republic March 2008
METAS Switzerland July 2008
NPL United kingdom June 2008
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 10
EUROMET.L-K7-2006_Final
Table 6: Changes in measurement schedule in group 2
Laboratory Country Date
INMETRO Brazil May 2007
CENAM Mexico June 2007
NIST USA July – August 2007
NRC Canada July 2008
A*Star–NMC Singapore August – September 2007
NIM China September – October 2007
NPLI India Oct 2007 – Feb.2008
NIMT Thailand March – April 2008
NPL United kingdom September 2008
METAS Switzerland June 2008
VNIIM Russia December 2008
2.5 Transport of the artefacts
Transport of the artefacts was critical in some cases outside Europe. In most cases fast courier services were used. The ATA carnet, which was used for the majority of laboratories outside EU, was mostly handled correctly. Sometimes it was not stamped correctly and once it was lost in Asia. But a substitutional ATA was issued and the problem was successfully solved. Special customs arrangements (temporary import) were necessary for some countries.
Packaging for the artefact was robust to protect the artefacts from being deformed or damaged during transit. The artefact was in an original NPL wooden box, which was put into a robust transport box (See Fig. 1). The outer transport box was wrapped in a cardboard box, which was replaced each time when it was worn out.
Fig. 1: Scale containers
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 11
EUROMET.L-K7-2006_Final
3 The artefact
3.1 Description of the artefact
At the Euromet TCL meeting in October 2005 in Bucharest it has been decided to measure a 100 mm quarz scale with 0.1 mm pitch. The artefact has been produced by NPL. Its basic purpose is to serve as a standard in precise industrial calibrations. The artefact is shown in Fig. 2.
Fig. 2: NPL line scale
The width of the scale lines is approx. 10 m. The scale is provided by two parallel horizontal lines at the beginning and at the end of the scale. The distance between those 2 lines is approx. 50 m. Some details of the scale can be seen in Fig. 3.
…….
Fig. 3: Details of the scale
Equal artefacts were used in both groups. The artefact for group 1 was marked with engraved letter “A” and the artefact for group 2 with “B”. The boxes were marked in the same manner with stickers.
Dimensions of the artefact are presented in Fig. 4.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 12
EUROMET.L-K7-2006_Final
2.35
mm
20
mm
115 mm
Fig. 4: Dimensions of the artefact
3.2 Fixing the artefact
The artefact was shipped without any special mounting fixtures. It was recommended to support the measurement objects at the Airy points (distance of x = 0.2113L from both ends), held only by their gravity forces. It was not allowed to use any type of glue or wax for mounting the scale. If additional clamping of the scale was required during measurement, e.g. because of a fast moving carriage, it was recommended to lightly pinch the scale on the sides at one of the Airy support points. If other support or clamping conditions were applied during measurement, it was the responsibility of the participant to refer his results to the Airy point support conditions.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 13
EUROMET.L-K7-2006_Final
4 Measurement instructions
4.1 Traceability
4.1.1 Length measurements should be traceable to the latest realisation of the metre as set out in the current “Mise en Pratique”.
4.1.2 Temperature measurements should be made using the International Temperature Scale of 1990 (ITS-90).
4.2 Measurand
Measurand was the distance between the centre line position of the reference line (position “0”) and the centre line position of the measured line (Fig. 5). To increase comparability of the results, all measurements were performed over the section between the two horizontal lines (at the beginning and at the end of the scale) with a width of approx. 50 µm. That is, it had to be tried to apply an effective slit height or CCD image window height of 50 µm for the analysis of measurements. If the effective height could not be set exactly to 50 µm, a value close to it should have been chosen.
0 1
m1 m2
m3
Measuring section of 50 µm
Fig. 5: Measurand (m1, m2, m3) and measuring section
Table 7: The lines (distances) that were measured:
Nominal lengths in mm
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
5 10 15 20 25 30 35 40 45 50
55 60 65 70 75 80 85 90 95 100
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 14
EUROMET.L-K7-2006_Final
Measurement conditions:
The positions of the lines had to be determined as the centre line positions1 of every line, while the scale was lying on the Airy points (see 3.2). The participants were asked to describe the way the position of the line was determined.
For alignment purposes of the graduation lines the upper horizontal lines at the beginning and at the end of the scale should be used.
The measured values had to be referred to the following reference conditions: - temperature of 20 °C (ITS-90), - pressure of 1013,25 hPa (1013,25 mbar).
If necessary, corrections had to be applied based upon the following parameters:
Quartz: - Thermal expansion coefficient: ................................. = 510-7 K-1 - Length compressibility: ............................................. = - 8.910-7 bar-1
4.3 Measurement instructions
4.3.1 The calibration had to be carried out as for a normal customer. The participants were free to choose their own method of measurement. However, under the assumption that the value of the measurand is a true property of the material measure of length, only one result for a measurand had to be given irrespective of the number of different measurement methods used. For each method applied, a complete description of the method had to be given. A detailed estimation of the measurement uncertainty according to the ISO Guide to the Expression of Uncertainty in Measurement (GUM) had to be supplied.
4.3.2 The measurements had to be reported for measuring conditions, given in 4.2.
4.3.3 Before calibration, the scale had to be inspected for damages. Any scratches, dirty spots or other damages had to be documented
4.3.4 The measurement results (appropriately corrected to the reference conditions) had to be reported using forms, given in the protocol.
4.4 Measurement uncertainty
The uncertainty of measurement had to be estimated according to the ISO Guide to the Expression of Uncertainty in Measurement. In order to achieve a better comparability, some possible influence parameters and notations were given. The participants were encouraged to use all known and significant influence parameters for their applied methods. The following list could have been used as an indication of possible influence parameters:
Possible contributions from line position sensing technique:
Eres Resolution of edge detection
sE Repeatability of edge detection
1 The key comparison guideline states, that the methods usually applied by the participants for calibrations should also be used within the comparison. Because different line center extraction algorithms will normally be used by the participants, it is essential that the different procedures are well described and that edge detection influences are accounted for in the uncertainty estimation. A possible edge detection algorithm e.g. is the arithmetic mean of left and right edge positions if those are explicitly measured (e.g. at 50% threshold) or the centroid of 2D image intensity data.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 15
EUROMET.L-K7-2006_Final
Edef Edge geometry influence (roughness, parallelism)
lpos Influence of adjustment of measurement line
lwin Influence of adjustment of measurement window or slit length
Efoc Influence of focal length variation
E Influence of detection light wavelength
Epol Influence of detection light polarization
Ecoh Influence of detection light coherence
Mag Microscope magnification (or other position deviation sensing device)
Enon Nonlinearities of position sensing technique
Ealig Microscope axis alignment
Ealg Influence of line edge detection algorithm, possible asymmetry of line profiles, line shape
Erev Influence of measurement in reversed orientation
Possible contributions from interferometric displacement measurement technique:
o vacuum wavelength of light source used for displacement measurement
nair Index of refraction of air2
tair Air temperature
pair Air pressure
RHair Air humidity
cCO2 Air CO2 concentration
lRes Interferometer resolution
lNL Interferometer nonlinearity (polarisation mixing, etc.)
lDP Interferometer dead path influences (temperature variation, etc.)
lMP Variation of measurement path in one orientation (normal, meander, random, ..)
lDrift Drift influence (forward, backward measurement)
lRev Influence of measurement in reversed orientation
lAi Errors due to Abbe offsets and pitch and yaw of translation stages
lSi Errors of scale alignment
lIi Cosine errors of interferometer alignment
Possible contributions from scale properties:
Z, Cr Linear coefficient of thermal expansion of scale material
ts = (ts - 20) is the difference of the scale temperature ts in °C during the measurement from the reference temperature of 20 °C
Z, Cr Linear coefficient of compressibility of scale material
h Flatness deviation of scale graduation surface
supp Influence of support conditions
2 If the index of refraction is determined by the parameter method according to Edlen, the updated version of the formula should be applied as published in: G. Bönsch, E. Potulski, Metrologia, 1998, 35, 133-139. The estimated combined standard uncertainty of the quoted formula itself is 1*10-8.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 16
EUROMET.L-K7-2006_Final
The deviations dL from nominal length had to be measured and expressed as a function of input quantities xi
)( ixfdL , (1)
The combined standard uncertainty uc(dl) is the quadratic sum of the standard uncertainties of the input quantities u(xi) each weighted by a sensitivity coefficient ci
ii
iiic x
dLcxucdLu
with,)()( 222 . (2)
The participants were required to report their measurement uncertainty budget in a prepared table (in Appendix A.2 of the technical protocol) with the format according to the scheme below. "Distrib." is the type of distribution of the input quantity (N=normal, R=rectangular, T=triangular, etc.), i is the number of degrees of freedom of u(xi).
Example scheme:
Input quantity xi Distrib. u(xi) i ci = dL / xi ui (dL) / nm
Edge detection reproduc. sE N 3 nm 10 1 3
Cosine error scale alignment R 140 µrad >100 - 10-8 L
... ... ... ... ... ...
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 17
EUROMET.L-K7-2006_Final
5 Measurement equipment and methods used by the participants
Detailed information on the equipment and method used by the participants is in Appendix 2. Short summary is presented in Table 8.
Table 8: Overview of instrumentation used by the participants
Laboratory Measuring instrument Line detection
MIRS
Zeiss ULM 01-600 C 1D measurement machine by using laser interferometer HP 5528A, fixed microscope, moving scale, hand driven
CCD microscope, in-house software
BEV SIP 3002 length measuring machine with a standard HP 5529A laser interferometer; fixed scale, moving microscope, hand driven
Incident light CCD-microscope; two parallel reference lines
CMI Interferometric comparator IK-1 (CMI design), fixed microscope, moving scale; motor driven
CCD microscope; in-house software
DZM-FSB
500 mm 1D machine, in-house design and construction, Renishaw ML 10 laser interferometer, fixed microscope, moving scale, hand driven
CCD microscope, in-house software
GUM
1000 mm 1D SIP measuring bench, laser interferometer HP-5528A, fixed microscope, moving scale, motor driven + precision piezo-electric actuator
CCD-microscope; two parallel reference lines
INM Longitudinal comparator, He – Ne frequency stabilized laser interferometer, fixed microscope, moving scale, motor driven
Optical microscope
LNMC Horizontal comparator IZA-7, longitudinal comparison, fixed scale, hand driven
2 microscopes
METAS
2D photomask measuring system 400 mm x 300 mm, differential two axis plane mirror interferometer (HP), fixed microscope, moving scale, motor driven, fully automated
CCD microscope, motorised focusing, in-house software
MIKES
MIKES‘ line scale interferometer, Michelson interferometer utilising a calibrated 633 nm Zeeman-stabilised He-Ne laser, dynamic method, moving microscope, motor driven
CCD microscope, synchronous data sampling
NCM Comparator, HP 5529A laser interferometer, fixed scale, moving microscope, hand driven
Photoelectric microscope
NML SIP horizontal measuring machine, Agilent 5519A laser, fixed microscope, moving scale, hand driven
Optical microscope
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 18
EUROMET.L-K7-2006_Final
Laboratory Measuring instrument Line detection
NPL
NPL 400mm range air-bearing stage (interferometrically monitored), two co-linear independent laser interferometers, (NPL differential Jamin type, HP Michelson), fixed microscope, moving scale (different conditions in 2006 and 2008), motor driven
NPL NanoVision image processing system
NSCIM
Horizontal comparator with Michelson dynamic laser interferometer, primary standard DETU 01-03-98, fixed scale, moving microscope, hand driven
Photoelectric microscope (PEM)
OMH 3 m Zeiss universal length measuring machine, HP 5528 laser interferometer, hand driven
CCD microscope, reference screen lines
PTB
PTB Nanometer Comparator, vacuum interferometer (iodine-stabilized, frequency doubled Nd:YAG laser, fixed microscope, moving scale, motor driven
PTB optical microscope with CCD camera, in-house software
SMU 1-D machine Abbe Zeiss (range up to 200 mm), laser interferometer HP 5529B, fixed microscope, moving scale, hand driven
Optical microscope
ZMDM Zeiss ULM 3000 1-D measuring machine, laser interferometer HP 5526 A, fixed scale, moving microscope, hand driven
Zeiss optical microscope
CEM Custom-built length comparator CEM-TEK 1200, laser-interferometer (Stabilized Laser Source HP 5517C)
In-house software
CENAM Optical microscope brand Leitz Libra 200, moving scale, hand driven
CCD microscope, manual edge observation, Micro/Measure Microscope Software
EIM Leitz universal measuring microscope, laser interferometer Renishaw, moving scale, hand driven
CCD microscope, analysis of digital images
INMETRO Optical CMM SIP Trioptic, laser interferometer, fixed microscope, moving scale, hand driven
CCD microscope
INRIM Moore Measuring Machine, laser interferometer HP 5518 fixed microscope, moving scale, motor driven
CCD microscope, in-house software
NIM NIM comparator, , He-Ne laser interferometer Optical-electronic microscope
dual slit line position detection
NIST
N1ST Line Scale lnterferometer (LSI) - heterodyne interferometer, He-Ne laser, fixed microscope, moving scale, motor driven
Scanning electro-optical line detector, in-house software
NMi-VSL BV
SIP 400 measuring machine, laser interferometer HP, fixed microscope, moving scale, hand driven
CCD microscope, in-house software
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 19
EUROMET.L-K7-2006_Final
Laboratory Measuring instrument Line detection
NPLI
Universal measuring machine SIP UMM – MUL -214 B, heterodyne laser interferometer HP 5529A, fixed microscope moving scale, hand driven
Optical microscope
NRC NRC 4-metre Line Scale Comparator, HP heterodyne laser interferometer, fixed scale, moving microscope, motor driven
CCD microscope, in-house software
A*Star - NMC
Laser line width measurement system (base with two working stages, motor drives, He-Ne laser interferometer), fixed microscope moving scale, motor driven
Photoelectric microscope with single slit
NIMT Line scale interferometer (stabilized He-Ne laser, fixed scale, moving edge sensor, motor (?) driven
Edge sensor using triple slits system
VNIIM
VNIIM comparator (carriage, laser polarization interferometer, refractometer) fixed scale, moving microscope, motor driven, measurements in dynamic mode
Microscope with laser diodes and photodiode
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 20
EUROMET.L-K7-2006_Final
6 Stability of the standards
The standards used for this comparison were made of quartz with expected good long term stability. However, no historical data were available, since the standards were new. The material properties (chapter 4) were not exactly determined, the values were obtained from general knowledge about the used material.
In order to check the long-term stability, measurements of two linking laboratories (METAS and NPL) were performed in the beginning of the project (year 2006) and close to the end of the project (2008).
The stability checks comprised the total length change during 2 years, as well as possible line centre shifts due to damages and dirtiness of single lines.
6.1 Stability of the total line scale length
Stability of the total line scale length (over 100 mm) is demonstrated in Figures 6 (Scale “A”, Group 1) and 7 (Scale “B”, Group 2). Deviations from nominal length, measured by the linking laboratories in years 2006 and 2008 are represented together with the standard uncertainties.
Stability - scale "A" (Group 1) - Difference in total scale length2006 and 2008 (METAS)
-405
-400
-395
-390
-385
-380
-375
-370
-365
2006 2008
Time
Dev
iati
on
s 20
06 a
nd
200
8 in
nm
Stability - scale "A" (Group 1) - Difference in total scale length2006 and 2008 (NPL)
-420
-410
-400
-390
-380
-370
-360
-350
-340
-330
2006 2008
Time
Dev
iati
on
s 20
06 a
nd
200
8 in
nm
Fig. 6: Measured deviations of the scale “A” in years 2006 and 2008
The result for scale A in Fig. 6 shows a slight drift in negative direction, which is however much smaller than the standard uncertainties of both laboratories. Therefore, no drift correction was applied to the measurement values.
Stability - scale "B" (Group 2) - Difference in total scale length2006 and 2008 (METAS)
-305
-300
-295
-290
-285
-280
-275
-270
-265
-260
-255
-250
2006 2008
Time
Dev
iati
on
s 20
06 a
nd
200
8 in
nm
Stability - scale "B" (Group 2) - Difference in total scale length2006 and 2008 (NPL)
-310
-300
-290
-280
-270
-260
-250
-240
-230
-220
-210
2006 2008
Time
Dev
iati
on
s 20
06 a
nd
200
8 in
nm
Fig. 7: Measured deviations of the scale “B” in years 2006 and 2008
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 21
EUROMET.L-K7-2006_Final
The result for scale B in Fig. 7 shows drifts in different directions. Because no systematic drift was identified, no corrections were applied to the measurement values to compensate for drift.
6.2 Stability of single line centres
The participating laboratories were instructed how to clean the scale before measurement and to report possible damages. The following damages and dirt spots were reported:
Fig. 8: MIKES reported small damage on the reference line (Scale “A”)
Dust on scale before and after cleaning
Spots on glass
Spots on a line
Fig. 9: Dirt and damages reported by PTB (scale “A”)
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 22
EUROMET.L-K7-2006_Final
Fig. 10: Scratches reported by NCM (scale “A”)
Fig. 11: Dirt and damages reported by A*Star–NMC Singapore (scale “B”)
After discussion between the pilot laboratory and both linking laboratories it was agreed that reported damages would not impact the majority of results and their uncertainties. The laboratories were also instructed in the technical protocol to perform the measurements as for their clients. So it was in their responsibility to consider disturbances in their uncertainty budget.
Single line centre shifts due to changes on lines can be evaluated from the diagrams in Figures 12 to 16. Results are presented in three different ways:
- as a comparison of measurements in 2006 and 2008 – Fig. 12 and 13
- as an absolute difference of two results (2008 – 2006) – Fig. 14
- as deviations of two measurements from calculated arithmetic mean (2006 and 2008) – Fig. 15 and 16
Stability - scale "A" (Group 1) - Comparison of results 2006 and 2008 (METAS)
-500
-400
-300
-200
-100
0
100
0,4
0,6
0,8 1 10 20 30 40 50 60 70 80 90 10
0
Nominal length in mm
Res
ult
s 20
06 a
nd
200
8 in
nm
METAS 2006
METAS 2008
Stability - scale "A" (Group 1) - comparison of results 2006 and 2008 (NPL)
-450
-400
-350
-300
-250
-200
-150
-100
-50
0
50
0 0,2
0,4
0,6
0,8 1 10 20 30 40 50 60 70 80 90 10
0
Nominal length in mm
Res
ult
s 20
06 a
nd
200
8 in
nm
NPL 2006
NPL 2008
Fig. 12: Comparison of measurements, performed by METAS and NPL in 2006 and in 2008 on scale “A”
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 23
EUROMET.L-K7-2006_Final
Stability - scale "B" (Group 2) - Comparison of results 2006 and 2008 (METAS)
-400
-350
-300
-250
-200
-150
-100
-50
0
50
0 0,2
0,4
0,6
0,8 1 10 20 30 40 50 60 70 80 90 10
0
Nominal length in mm
Res
ult
s 20
06 a
nd
200
8 in
nm
METAS 2006
METAS 2008
Stability - scale "B" (Group 2) - comparison of results 2006 and 2008 (NPL)
-350
-300
-250
-200
-150
-100
-50
0
50
0 0,2
0,4
0,6
0,8 1 10 20 30 40 50 60 70 80 90 10
0
Nominal length in mm
Res
ult
s 20
06 a
nd
200
8 in
nm
NPL 2006
NPL 2008
Fig. 13: Comparison of measurements, performed by METAS and NPL in 2006 and in 2008 on scale “B”
Stability - scale "A" (Group 1)
-25
-20
-15
-10
-5
0
5
10
0 0,2
0,4
0,6
0,8 1 10 20 30 40 50 60 70 80 90 10
0
Nominal length in mm
Dev
iati
on
200
8 -
2006
in n
m
METAS
NPL
Stability - scale "B" (Group 2)
-20
-15
-10
-5
0
5
10
15
0 0,2
0,4
0,6
0,8 1 10 20 30 40 50 60 70 80 90 10
0
Nominal length in mm
Dev
iati
on
in n
m
METAS
NPL
Fig. 14: Absolute difference of the results from 2006 and 2008 for scales “A” and “B”
Stability - scale "A" (Group 1) - deviation from mean (METAS)
-15
-10
-5
0
5
10
15
0 0,2
0,4
0,6
0,8 1 10 20 30 40 50 60 70 80 90 10
0
Nominal length in mm
Dev
iati
on
fro
m m
ean
in
nm
dev 2006
dev 2008
u
-u
Stability - scale "A" (Group 1) - deviation from mean (NPL)
-40
-30
-20
-10
0
10
20
30
40
0 0,2
0,4
0,6
0,8 1 10 20 30 40 50 60 70 80 90 10
0
Nominal length in mm
Dev
iati
on
fro
m m
ean
in
nm
dev 2006
dev 2008
u
-u
Fig. 15: Deviations of 2006 and 2008 measurements from the mean value (scale A) with indicated standard uncertainties
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 24
EUROMET.L-K7-2006_Final
Stability - scale "B" (Group 2) - deviation from mean (METAS)
-15
-10
-5
0
5
10
15
0 0,2
0,4
0,6
0,8 1 10 20 30 40 50 60 70 80 90 10
0
Nominal length in mm
Dev
iati
on
fro
m m
ean
in
nm
dev 2006
dev 2008
u
-u
Stability - scale "B" (Group 2) - deviation from mean (NPL)
-40
-30
-20
-10
0
10
20
30
40
0 0,2
0,4
0,6
0,8 1 10 20 30 40 50 60 70 80 90 10
0
Nominal length in mm
Dev
iati
on
fro
m m
ean
in
nm
dev 2006
dev 2008
u
-u
Fig. 16: Deviations of 2006 and 2008 measurements from the mean value (scale B) with indicated standard uncertainties
Figures 12 and 13 showing the deviations from the nominal values are not very useful, since the differences can hardly be seen. Fig. 14 shows for scale “A” the biggest difference of 23 nm measured by NPL, while the biggest difference in METAS results is 6 nm. For scale “B”, the biggest difference measured by NPL was 15 nm, while the biggest difference measured by METAS was 12 nm. Fig. 15 and 16 show, that all measured differences were within the standard uncertainty boundaries.
6.3 Impact of line scale instability
Although no systematic change in line scale length was detected, there are possible influences on measured results by random influences like dirt, damages, different cleaning approaches etc. These influences were taken into account by adding an additional uncertainty component before calculating the reference values and their uncertainties and consistency of the results (En). This uncertainty component derived from the stability measurements made by the linking laboratories. They measured the scale at the beginning and at the end of the comparison. A time dependent contribution (stability uncertainty) has to be added to a calibration value because the value is used at a time considerably later than the measurement. So adding the stability contribution makes the calibration valid over the full time span of the comparison (which lasted more than 2 years).
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 25
EUROMET.L-K7-2006_Final
7 Analysis of the results
The reference values (xref) were calculated as the weighted mean of all measurements (xi), the weight factors being u-2(xi). For each measurement point the reference value was calculated. In the second step, for the calculation of reference values from the largest consistent subset of measurements, some of the values with 1nE were omitted one by one, beginning with the
largest nE , until the Birge criterion was met.
Reference value:
)(
)(
1
2
1
2
i
n
i
ii
n
iref
xu
xxux
(3)
Combined standard uncertainty:
)(
1)(
1
2i
n
i
refc
xu
xu
(4)
En-value:
2;)()( 22
k
xuxuk
xxE
reflab
reflabn (5)
The Birge ratio:
intu
uR ext
B (6)
)()1(
)(/)(
1
2
2
1
i
n
i
i
n
irefi
ext
xun
xuxx
u
(7)
)(/11
2int i
n
i
xuu
(8)
The Birge criterion:
1
81
nRB
(9)
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 26
EUROMET.L-K7-2006_Final
8 Measurement results
8.1 Measurement results - Group 1
Measurement results of Group 1 are presented in Table 9. The first set of NPL measurements and the last set of METAS measurements were used only for scale stability evaluation. They were excluded from final evaluation of the participants.
Table 9: Measurement results – Group 1
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 27
EUROMET.L-K7-2006_Final
Figure 17 presents measurement results (deviations from nominal values) for measurement ranges 0,1 mm to 1 mm and 1 mm to 100 mm.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 28
EUROMET.L-K7-2006_Final
-250
-200
-150
-100
-50
0
50
100
0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1
Rep
orte
d de
viat
ions
measurement points
Measured values 0,1 mm to 1 mm (excl. LNMC, NML)
NCM
MIRS
OMH
BEV
NPL
SMU
PTB
MIKES
INM
DZM
CMI
GUM
METAS
NSCIM
ZMDM
-800
-600
-400
-200
0
200
400
1 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
Rep
orte
d de
viat
ions
measurement points
Measured values 1 mm to 100 mm (excl. ZMDM, LNMC, NML)
NCM
MIRS
OMH
BEV
NPL
SMU
PTB
MIKES
INM
DZM
CMI
GUM
METAS
NSCIM
Figure 17: Deviations from nominal values - group 1
Graphical illustration of deviations from the reference values, which were calculated separately for group 1 (without linking the groups) are presented in Appendix 1.
8.2 Measurement results – Group 2
Measurement results of Group 2 are presented in Table 10. The first set of NPL measurements and the last set of METAS measurements were used only for scale stability evaluation. They were excluded from final evaluation of the participants.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 29
EUROMET.L-K7-2006_Final
Table 10: Measurement results – Group 2
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 30
EUROMET.L-K7-2006_Final
Figure 18 presents measurement results (deviations from nominal values) for measurement ranges 0,1 mm to 1 mm and 1 mm to 100 mm.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 31
EUROMET.L-K7-2006_Final
-150
-100
-50
0
50
100
0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1
rep
ort
ed
de
via
tio
ns
/ n
m
measurement points / mm
Measured values 0,1 mm to 1 mm
METAS
NIST
NMI-VSL
VNIIM
NPL
INRIM
CEM
NRC
NIMT
EIM
INMETRO
CENAM
A*Star
NIM
NPLI
-800
-700
-600
-500
-400
-300
-200
-100
0
100
200
1 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
rep
ort
ed
de
via
tio
ns
/ n
m
measurement points / mm
Measured values 1 mm to 100 mm
METAS
NIST
NMI-VSL
VNIIM
NPL
INRIM
CEM
NRC
NIMT
EIM
INMETRO
CENAM
A*Star
NIM
NPLI
Figure 18: Deviations from nominal values – group 2
Graphical illustration of deviations from the reference values, which were calculated separately for group 2 (without linking the groups) are presented in Appendix 1.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 32
EUROMET.L-K7-2006_Final
8.3 Correction of the original results
After issuing draft A report, one laboratory (SMU from Group 1) requested to correct its results due to wrong thermal expansion correction, and one laboratory (A*Star from group 2) requested to correct its uncertainty budget due to obvious error in calculating standard uncertainties of input quantities.
After the two laboratories had sent their technical explanations, the requested changes were approved, because clear mistakes were shown. The explanations are attached below:
SMU Slovakia:
“SMU used Abbe Zeiss length measuring machine combined with HP 5529B laser interferometer. The computer was working permanently during the measurement period (December 2006), i.e. it ran approximately 3 weeks without switch off. The external sensors have been used and the actualized values of the line measure temperature and atmosphere parameters were being inserted into the system before each measurement cycle (10 lines within each measurement cycle). Of course, the thermal dilatation coefficient (in the form x.xxx ppm/°C) was inserted once at the beginning of SMU measurements and thus its value remained untouched during the whole measurement period.
The most frequently used values while measuring steel or glass artifacts are therefore 11.500 ppm/°C or 8.500 ppm/°C respectively. Unfortunately, only after receipt of the first draft I realized that the value I saw displayed was 5.000 ppm/°C (instead of 0.500 ppm/°C). Nevertheless, I must admit that after almost 3 years I am not quite sure (i.e. 100%) what I actually saw on the display, but it is much more probable that 0.5 ppm was not displayed there.
The average temperature corresponding to the larger nominal lengths was 19,63 °C, therefore the temperature compensation according to the value of 5x10-6 K-1 was 10 times larger and hence it shifted the measured values towards the nominal ones (of course, just those lying below the nominal, but in fact it seems that for entirely all larger values it is the case). For example, in the corrected set of SMU results, the deviation corresponding to 100 mm has changed from –84 nm to –250 nm, etc.
Paradoxically, I always kept in mind that the thermal expansion coefficient of quartz is low and thus I did not care for the temperature as usually, not in terms of the temperature measurement itself, but in terms of the communication with the air condition centre. The air conditioning system of the institute is old, expensive and not as effective as in many other institutes, therefore we have always problem to keep the temperature close to 20 °C, even if it is relatively stable. It was December and thus the temperature in lab was below 20 °C. Just three weeks I had for measurements (we have obliged holiday from Christmas to New Year at the SMU) and there was not enough time to wait until the ambient temperature is closer to 20 °C.
The microscope magnification I used was 125x and the line edge was of high quality; according to my experience, I can hardly believe that I could be wrong by roughly 0.5 �m as it looks from the first submitted set of measurements. I am aware that some people could take the explanation given above as the clear speculation, because of the subjective feeling of “what I saw on the display 2.5 years ago” can be hardly taken as the rigid proof in this case.”
A*Star Singapore:
“A*Star-NMC-SG has requested a change in two uncertainty components being the errors due to horizontal and vertical Abbe offsets. The initial values of the two components were given at 2 sigma level with rectangular distribution and should be divided by SQRT(3) for the conversion to standard uncertainty. The miss-calculations were corrected and updated in the uncertainty budget table.”
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 33
EUROMET.L-K7-2006_Final
8.4 Link between the groups
The results in the previous chapter were shown for each group separately. This chapter is presenting the results after linking the groups by using Bayesian statistics 2.
In order to make statistical formulas more clear, Group 1 is indexed with “A” and Group 2 with “B” as shown in Figure 19. Index “C” is assigned to the intersection of sets “A” and “B” , which contains 2 elements (linking laboratories NPL and METAS).
Figure 19: Group 1 and Group 2 are presented by 2 sets of elements “A” and “B”
The results are calculated by the following formulas:
Reference values:
221
A cab
cSbSy
(10)
221
B cab
aScSy
(11)
Standard uncertainties of the reference values:
2A )(cab
byu
(12)
2B)(cab
ayu
(13)
Covariance:
2BA ),(cab
cyyu
(14)
Where:
BABA ,B,A
2,B
2,A
2,B
2
\ ,A2 ,
1
i iiii
i
i i xxuxuxu
xu
xua (15)
BAAB ,B,A
2,B
2,A
2,A
2
\ ,B2 ,
1
i iiii
i
i i xxuxuxu
xu
xub (16)
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 34
EUROMET.L-K7-2006_Final
BA ,B,A
2,B
2,A
2,B,A
,
,
i iiii
ii
xxuxuxu
xxuc (17)
BABA ,B,A2
,B2
,A2
,B,B,A,A,B2
\ ,A2
,A1 ,
,
i iiii
iiiii
i i
i
xxuxuxu
xxxuxxu
xu
xS (18)
BABA ,B,A2
,B2
,A2
,A,B,A,B,A2
\ ,B2
,B2 ,
,
i iiii
iiiii
i i
i
xxuxuxu
xxxuxxu
xu
xS (19)
Values x are reported results of participating laboratories (xA,i for Group 1 and xB,i for Group 2), while u(x) are reported standard uncertainties.
8.5 Intercomparison results for linked groups considering line scale stability
The results are shown in Table 12. Linking laboratories are in the middle of the table and are marked with green colour. Above them are laboratories from Group 1, while the laboratories from Group 2 are in the bottom part of the table.
The results given in Table 12 were calculated from the largest consistent subset. This subset was created by eliminating laboratories with the greatest En values until the Birge criterion (Ch. 7) was met and Chi-test passed. En values 1nE are marked with yellow colour. Eliminated
laboratories for single measurement points are shown in Table 11.
Table 11: Laboratories that were excluded from calculation of the reference values:
Measuring point (mm)
No. of excl. labs
Excluded laboratories
Measuring point (mm)
No. of excl. labs
Excluded laboratories
0,6 1 NSCIM-UA 55 2 ZMDM-SR, NSCIM-UA
5 1 ZMDM-SR 60 3 ZMDM-SR, NSCIM-UA, NML-IE
10 1 ZMDM-SR 65 2 ZMDM-SR, NSCIM-UA
15 2 ZMDM-SR, NSCIM-UA 70 2 ZMDM-SR, NSCIM-UA
20 2 ZMDM-SR, NSCIM-UA 75 3 ZMDM-SR, NSCIM-UA, NML-IE
25 2 ZMDM-SR, NSCIM-UA 80 3 ZMDM-SR, NSCIM-UA, NML-IE
30 2 ZMDM-SR, NSCIM-UA 85 5
ZMDM-SR, NSCIM-UA, NML-IE, OMH-HU, BEV-AT, NIM-CN
35 2 ZMDM-SR, NSCIM-UA 90 5
ZMDM-SR, NSCIM-UA, NML-IE, SMU-SK, INM-RO
40 2 ZMDM-SR, NSCIM-UA 95 4
ZMDM-SR, NSCIM-UA, NML-IE, SMU-SK
45 2 ZMDM-SR, NSCIM-UA 100 5
ZMDM-SR, NSCIM-UA, NML-IE, BEV-AT, INM-RO, NIM-CN
50 2 ZMDM-SR, NSCIM-UA
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 35
EUROMET.L-K7-2006_Final
Table 12: Reference values and uncertainties for all lines of scale A and scale B. Additionally, the deviations and the consistency with the reference values (En) are given for each participant
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 36
EUROMET.L-K7-2006_Final
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 37
EUROMET.L-K7-2006_Final
Graphical presentations of measured results indicated as deviations from calculated reference values follow in the next 30 diagrams. The dotted lines indicate the uncertainties of the reference values. Uncertainty bars in the diagrams represent expanded uncertainty U (k = 2).
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 0,1 mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 38
EUROMET.L-K7-2006_Final
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 0,2 mm
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 0,3 mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 39
EUROMET.L-K7-2006_Final
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 0,4 mm
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 0,5 mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 40
EUROMET.L-K7-2006_Final
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 0,6 mm
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 0,7 mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 41
EUROMET.L-K7-2006_Final
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 0,8 mm
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 0,9 mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 42
EUROMET.L-K7-2006_Final
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 1 mm
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 5 mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 43
EUROMET.L-K7-2006_Final
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 10 mm
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 15 mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 44
EUROMET.L-K7-2006_Final
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 20 mm
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 25 mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 45
EUROMET.L-K7-2006_Final
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 30 mm
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 35 mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 46
EUROMET.L-K7-2006_Final
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 40 mm
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 45 mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 47
EUROMET.L-K7-2006_Final
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 50 mm
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 55 mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 48
EUROMET.L-K7-2006_Final
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 60 mm
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 65 mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 49
EUROMET.L-K7-2006_Final
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 70 mm
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 75 mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 50
EUROMET.L-K7-2006_Final
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 80 mm
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 85 mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 51
EUROMET.L-K7-2006_Final
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 90 mm
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 95 mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 52
EUROMET.L-K7-2006_Final
‐400,0
‐300,0
‐200,0
‐100,0
0,0
100,0
200,0
300,0
400,0
MIRS‐SI
OMH‐HU
BEV
‐AT
SMU‐SK
PTB
‐DE
GUM‐PL
MIKES‐FI
LNMC‐LV
NML‐IE
NCM‐BG
INM‐RO
ZMDM‐SR
DZM
‐HR
NSCIM
‐UA
CMI‐CZ
NPL‐GB
METAS‐CH
NPL‐GB
METAS‐CH
EIM‐GR
INRIM
‐IT
Nmi‐VSL‐NL
CEM
‐ES
INMETRO‐BR
CEN
AM‐M
X
NIST‐US
NRC‐CA
A*Star‐NMC‐SG
NIM
‐CN
NPLI‐IN
NIM
T‐TH
A
VNIIM
‐RU
Deviations / nm
Deviations from the reference value at nominal length 100 mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 53
EUROMET.L-K7-2006_Final
9 Conclusions
The intention of this comparison was to determine and to document capabilities of the participating NMIs to carry out line scale calibrations on high quality line scales produced for industrial purposes. The line scales used for the comparison were designed and produced by NPL in UK. Two scales of same design and very similar quality were kindly donated by NPL.
The idea for the comparison arised at the Euramet TCL meeting in October 2005. The comparison started in July 2006 and the last measurement was performed in December 2008. Originally, 31 NMIs expressed interest for participating in the comparison. During the comparison, two laboratories decided not to perform measurements due to technical reasons and one new laboratory was approved to take part. At the end 30 laboratories reported their results.
Participating laboratories were divided into 2 groups in accordance with their geographical position (in order to minimize travel times and expenses for the transportation of the standards). Linking laboratories between the groups were chosen among participants in Nano3 project (NPL and METAS).
Although the standards traveled through a large number of laboratories, no significant damages were noticed. Some laboratories reported some dirt and scratches, but no significant influence on the results were indicated. The comparison ran quite well within the schedule in spite of some customs problems. Changes in the schedule are indicated in chapter 2.4.
Results were evaluated for each group separately and also after linking groups by using Bayesian statistics [2]. The performance of the participants was evaluated by using En value as the acceptance criterion. The reference value was calculated as the weighted mean of reported results for each measuring point. The Birge criterion and Chi-test were used for approving calculated reference values.
In the conclusion it can be summarised that the comparison was successful and has shown realistic picture about calibration and measurement capabilities of participating laboratories.
10 Acknowledgment
The pilot laboratory would like to thank all involved experts and other staff from the participating NMIs for their co-operation and performed technical and formal work. Special thanks should be addressed to the linking laboratories NPL-UK and METAS-CH for offering technical support through extensive discussions about the analysis of the results.
However, the comparison would not be possible without the generous support of the producer and owner of the line scales – NPL from UK. At this occasion the pilot laboratory is expressing very special thanks to NPL and to Dr. Michael McCarthy personally.
11 References
[1] M. G. Cox. The evaluation of key comparison data, Metrologia 39, 589–595 (2002)
[2] M. Krystek; personal communication
[3] H. Bose. Nano 3 – Line Scale Standards; WGDM-7 Preliminary comparison on nanometrology, Final report (2003)
[4] R. Thalmann. EUROMET 677- Steel Tape Measures, Final report (2004)
[5] Guidelines for CIPM key comparisons, (2003), http://www.bipm.org/utils/en/pdf/guidelines.pdf
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 54
EUROMET.L-K7-2006_Final
Appendix 1: Intercomparison results for the case of separate treatment of groups 1 and 2 (without link)
1.1 Group 1 – deviations from reference value
Graphical presentation of measurement results follows in 30 diagrams. Uncertainty bars in the diagrams for single measuring points represent standard uncertainty u (k = 1).
-400
-300
-200
-100
0
100
200
300
400
MIR
S-S
I
ME
TAS
-CH
OM
H-H
U
BE
V-A
T
SM
U-S
K
PTB
-DE
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-CZ
NP
L-G
B
nmResults Group 1; 0,1 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
400
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 0,2 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
400
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 0,3 mm
dev
xref
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 55
EUROMET.L-K7-2006_Final
-400
-300
-200
-100
0
100
200
300
400
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 0,4 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
400
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nm
Results Group 1; 0,5 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
400
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 0,6 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
400
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LN
MC
-LV
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 0,7 mm
dev
xref
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 56
EUROMET.L-K7-2006_Final
-400
-300
-200
-100
0
100
200
300
400
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 0,8 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
400
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LN
MC
-LV
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 0,9 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
400
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-I
E
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 1 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
400
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 5 mm
dev
xref
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 57
EUROMET.L-K7-2006_Final
-400
-300
-200
-100
0
100
200
300
400
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-I
E
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 10 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
400
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-I
E
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 15 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
400
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nm
Results Group 1; 20 mm
dev
xref
-400
-300
-200
-100
0
100
200
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 25 mm
dev
xref
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 58
EUROMET.L-K7-2006_Final
-400
-300
-200
-100
0
100
200
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 30 mm
dev
xref
-400
-300
-200
-100
0
100
200
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LN
MC
-LV
NM
L-I
E
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 35 mm
dev
xref
-500
-400
-300
-200
-100
0
100
200
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LN
MC
-LV
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 40 mm
dev
xref
-500
-400
-300
-200
-100
0
100
200
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 45 mm
dev
xref
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 59
EUROMET.L-K7-2006_Final
-400
-300
-200
-100
0
100
200
300
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LN
MC
-LV
NM
L-I
E
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 50 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 55 mm
dev
xref
-500
-400
-300
-200
-100
0
100
200
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 60 mm
dev
xref
-500
-400
-300
-200
-100
0
100
200
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LN
MC
-LV
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 65 mm
dev
xref
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 60
EUROMET.L-K7-2006_Final
-500
-400
-300
-200
-100
0
100
200
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 70 mm
dev
xref
-500
-400
-300
-200
-100
0
100
200
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 75 mm
dev
xref
-600
-500
-400
-300
-200
-100
0
100
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 80 mm
dev
xref
-700
-600
-500
-400
-300
-200
-100
0
100
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 85 mm
dev
xref
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 61
EUROMET.L-K7-2006_Final
-800
-700
-600
-500
-400
-300
-200
-100
0
100
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 90 mm
dev
xref
-800
-700
-600
-500
-400
-300
-200
-100
0
100
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 95 mm
dev
xref
-800
-700
-600
-500
-400
-300
-200
-100
0
100
MIR
S-S
I
ME
TA
S-C
H
OM
H-H
U
BE
V-A
T
SM
U-S
K
PT
B-D
E
GU
M-P
L
MIK
ES
-FI
LNM
C-L
V
NM
L-IE
NC
M-B
G
INM
-RO
ZM
DM
-SR
DZ
M-H
R
NS
CIM
-UA
CM
I-C
Z
NP
L-G
B
nmResults Group 1; 100 mm
dev
xref
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 62
EUROMET.L-K7-2006_Final
1.2 Group 1 - Calculated values
The results given in Table 1.2 were calculated from the largest consistent subset. This subset was created by eliminating laboratories with the greatest En values until the Birge criterion (Ch. 7) was met. En values 1nE are marked with yellow colour. Eliminated laboratories for single
measurement points are shown in Table 1.1.
Table 1.1: Laboratories that were excluded from calculation of the reference values
Measuring point (mm)
No. of excl. labs
Excluded laboratories
Measuring point (mm)
No. of excl. labs
Excluded laboratories
0,6 1 NSCIM-UA 55 2 ZMDM-SR, NSCIM-UA
5 1 ZMDM-SR 60 3 ZMDM-SR, NSCIM-UA, NML-IE
10 1 ZMDM-SR 65 2 ZMDM-SR, NSCIM-UA
15 2 ZMDM-SR, NSCIM-UA 70 2 ZMDM-SR, NSCIM-UA
20 2 ZMDM-SR, NSCIM-UA 75 3 ZMDM-SR, NSCIM-UA, NML-IE
25 2 ZMDM-SR, NSCIM-UA 80 3 ZMDM-SR, NSCIM-UA, NML-IE
30 2 ZMDM-SR, NSCIM-UA 85 5
ZMDM-SR, NSCIM-UA, NML-IE, SMU-SK, INM-RO
35 2 ZMDM-SR, NSCIM-UA 90 5
ZMDM-SR, NSCIM-UA, NML-IE, SMU-SK, INM-RO
40 2 ZMDM-SR, NSCIM-UA 95 4
ZMDM-SR, NSCIM-UA, NML-IE, SMU-SK
45 2 ZMDM-SR, NSCIM-UA 100 5
ZMDM-SR, NSCIM-UA, NML-IE, SMU-SK, INM-RO
50 2 ZMDM-SR, NSCIM-UA
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 63
EUROMET.L-K7-2006_Final
Table 1.2: Results calculated from the given values – Group 1
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 64
EUROMET.L-K7-2006_Final
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 65
EUROMET.L-K7-2006_Final
1.3 Group 2 – deviations from reference value
Graphical presentation of measurement results follows in 30 diagrams. Uncertainty bars in the diagrams for single measuring points represent standard uncertainty u (k = 1).
-400
-300
-200
-100
0
100
200
300
400
ME
TA
S-C
H
EIM
-GR
INR
IM-I
T
Nm
i-V
SL
-NL
CE
M-E
S
INM
ET
RO
-BR
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-
IN
NIM
T-T
HA
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 0,1 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
400
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 0,2 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
400
ME
TA
S-C
H
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ET
RO
-BR
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-I
N
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nm
Results Group 2; 0,3 mm
dev
xref
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 66
EUROMET.L-K7-2006_Final
-400
-300
-200
-100
0
100
200
300
400
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ET
RO
-BR
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-IN
NIM
T-T
HA
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 0,4 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
400
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 0,5 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
400
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 0,6 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
400
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 0,7 mm
dev
xref
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 67
EUROMET.L-K7-2006_Final
-400
-300
-200
-100
0
100
200
300
400
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 0,8 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
400
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 0,9 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
400
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 1 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
400
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 5 mm
dev
xref
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 68
EUROMET.L-K7-2006_Final
-400
-300
-200
-100
0
100
200
300
400
ME
TA
S-C
H
EIM
-GR
INR
IM-I
T
Nm
i-VS
L-N
L
CE
M-E
S
INM
ET
RO
-BR
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-I
N
NIM
T-T
HA
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 10 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
400
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 15 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
400
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 20 mm
dev
xref
-400
-300
-200
-100
0
100
200
ME
TA
S-C
H
EIM
-GR
INR
IM-I
T
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-
SG
NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 25 mm
dev
xref
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 69
EUROMET.L-K7-2006_Final
-400
-300
-200
-100
0
100
200
ME
TA
S-C
H
EIM
-GR
INR
IM-I
T
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-
SG
NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 30 mm
dev
xref
-400
-300
-200
-100
0
100
200
ME
TA
S-C
H
EIM
-GR
INR
IM-I
T
Nm
i-VS
L-N
L
CE
M-E
S
INM
ET
RO
-BR
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-
SG
NIM
-CN
NP
LI-I
N
NIM
T-T
HA
NP
L-G
B
VN
IIM-R
U
nm
Results Group 2; 35 mm
dev
xref
-500
-400
-300
-200
-100
0
100
200
ME
TAS
-CH
EIM
-GR
INR
IM-I
T
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-
SG
NIM
-CN
NP
LI-I
N
NIM
T-T
HA
NP
L-G
B
VN
IIM-R
U
nm
Results Group 2; 40 mm
dev
xref
-500
-400
-300
-200
-100
0
100
200
ME
TAS
-CH
EIM
-GR
INR
IM-I
T
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-
SG
NIM
-CN
NP
LI-I
N
NIM
T-T
HA
NP
L-G
B
VN
IIM-R
U
nm
Results Group 2; 45 mm
dev
xref
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 70
EUROMET.L-K7-2006_Final
-400
-300
-200
-100
0
100
200
300
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-
SG NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 50 mm
dev
xref
-400
-300
-200
-100
0
100
200
300
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nm
Results Group 2; 55 mm
dev
xref
-500
-400
-300
-200
-100
0
100
200
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nm
Results Group 2; 60 mm
dev
xref
-500
-400
-300
-200
-100
0
100
200
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nm
Results Group 2; 65 mm
dev
xref
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 71
EUROMET.L-K7-2006_Final
-500
-400
-300
-200
-100
0
100
200
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nm
Results Group 2; 70 mm
dev
xref
-500
-400
-300
-200
-100
0
100
200
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 75 mm
dev
xref
-600
-500
-400
-300
-200
-100
0
100
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 80 mm
dev
xref
-700
-600
-500
-400
-300
-200
-100
0
100
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 85 mm
dev
xref
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 72
EUROMET.L-K7-2006_Final
-800
-700
-600
-500
-400
-300
-200
-100
0
100
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 90 mm
dev
xref
-800
-700
-600
-500
-400
-300
-200
-100
0
100
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 95 mm
dev
xref
-800
-700
-600
-500
-400
-300
-200
-100
0
100
ME
TAS
-CH
EIM
-GR
INR
IM-IT
Nm
i-VS
L-N
L
CE
M-E
S
INM
ETR
O-B
R
CE
NA
M-M
X
NIS
T-U
S
NR
C-C
A
A*S
tar-
NM
C-S
G NIM
-CN
NP
LI-IN
NIM
T-TH
A
NP
L-G
B
VN
IIM-R
U
nmResults Group 2; 100 mm
dev
xref
1.4 Group 2 - Calculated values
The results given in Table 1.4 were calculated from the largest consistent subset. This subset was created by eliminating laboratories with the greatest En values until the Birge criterion (Ch. 7) was met. En values 1nE are marked with yellow colour. Eliminated laboratories for single
measurement points are shown in Table 1.3.
Table 1.3: Laboratories that were excluded from calculation of the reference values
Measuring point (mm)
No. of excluded laboratories
Excluded laboratories
100 1 NIM-CN
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 73
EUROMET.L-K7-2006_Final
Table 1.4: Results calculated from the given values – Group 2
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 74
EUROMET.L-K7-2006_Final
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 75
EUROMET.L-K7-2006_Final-Appendix
Appendix 2: Descriptions of measurement setups and methods as reported by the participants
GROUP 1
BEV Austria
SIP 3002 length measuring machine with a standard HP 5529A laser interferometer only (i.e. no internal scale.) A standard linear interferometer, an Agilent 10751D air sensor, and 3 material temperature sensors were used. All systems calibrated by the BEV. The instrument is equipped with an incident light CCD-microscope. A special holder was constructed to ensure that the scale axis and the measurement axis of the laser interferometer coincide to within 0.1 mm.
For these measurements a 100 NA 0.55 objective with LED illumination was used. The image of the scale marks (as seen on a video monitor) were placed between two fixed lines (produced by a line generator) by manually moving the carriage together with the microscope. This visual technique together with residual straightness errors causes the most important uncertainty contribution.
To evaluate this uncertainty contributions measurements were performed by 4 different operators and on 5 diverse locations (orientation) on the machine. The stated results are the mean of this 5 measurements which are themselves the means of the observer-specific values. Because of instrumental boundary conditions the actual measurements were performed in two steps: 1 mm all 0.1 mm and 100 mm all 5 mm, respectively. The uncertainty for this two configurations is significantly different mainly of relocking effects.
Thermal drift was compensated by a time symmetrical measurement scheme. The scale temperature was estimated by measuring the table temperature some 5 cm away from the artefact. The scale support was near the end of the artefact (i.e. not at the Airy points).
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 76
EUROMET.L-K7-2006_Final-Appendix
CMI Czech Republic
CMI uses interferometric comparator IK-1 (CMI design) with static CCD microscope (objective 20x, ~130nm/pixel), transmission illumination by adjustable green or red LED, adjustable moving stage with linescale and interferometer retro-reflector.
Image processing:
line image is parallel to CCD rows: all pixels from selected range of one row are first added, then positions of both edges is found by linear interpolation of row sums, average of this two edge positions is taken as line position.
The stage automatically moves to predefined positions with about ±4µm precision (each line always close to the centre of the field of view). The readings of interferometer and CCD residual deviation are taken in synchronization. Resolution, noise and vibrations are bellow 10 nm. The stage returns to reference (zero) line after each movement.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 77
EUROMET.L-K7-2006_Final-Appendix
DZM Croatia
The measuring range of the device is 500 mm and it is primarily intended for the calibration of precise line scales. Device is fully constructed and produced at our Laboratory.
Stage movement is done manually.
The sighting process is done by means of a microscope with a digital CCD camera Olympus DP 70 with 12, 5 Megapixels with objective magnification of 50X.
The measuring system used is the laser interferometer (Reinshaw ML 10). The basis of the Renishaw Laser Interferometer system is HeNe Laser operating at a wavelength of 0,663 μm. The EC 10 Environmental Compensation unit automatically measures the three critical environmental parameters (temperature, pressure and relative humidity) and passes the data to the CS 10 Control Unit of which will compensate for any resultant change in wavelength.
The image processing is static, which means that it is necessary to process images and not the "live" signal of the display provided by a CCD camera. The software solution functions in such a way that all the pixels of a certain image are transmitted into a black&white combination and then the position of the line centre is calculated by arithmetic algorithms. The software solution provides the exact position of the line centre in pixels. In order to convert the values in pixels into the length values, it is necessary to calibrate the pixels size, i.e. to find out the length amount of every pixel. Calibration of the pixels is done absolutely by taking two images of the same line, which are shifted by the distance read on the laser interferometer.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 78
EUROMET.L-K7-2006_Final-Appendix
GUM Poland
The 1-D measuring bench (SIP) with total length of one meter is the base of the line scale measurement set-up. The set-up is located in the air-conditioned lab, housed and mounted on the concrete plate supported by the pneumatic vibroisolation system. The scale lines are detected using the microscope and the CCD-camera, mounted on the fixed column. The lines are observed on the monitor equipped with two parallel lines generated electronically. The distance of the generated lines can be adjusted for the artifact to be measured. The scale lines are aligned between this two lines by eye in a symmetrical manner with precision piezo-electric actuator. The line spacing is measured with the laser interferometer HP-5528A. The line scale was placed on the moving, remote controlled carriage. The linear reflector of the interferometer is mounted on the carriage. The measurement data are collected with PC computer through the HP-IB interface.
The line scale was supported, during the measurement, at the Airy points using gauge blocks. The measurements were carried out in normal and reversed orientation (0° and 180°). There was applied the image window height close to 50 µm, the microscope with 20x objective and total magnification of 770x for the analysis of measurements.
The environmental parameters (material and air temperature) were measured by interferometer sensors. Humidity was measured with thermo-hygrometer LB-706. The refractive index of air was calculated from the modified Edlen formula [G. Bönsch and E. Potulski, Metrologia, 1998, 35] with sensors corrections.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 79
EUROMET.L-K7-2006_Final-Appendix
INM Romania
Instrument:
- longitudinal comparator, equipped with an optical microscope and a laser interferometer with laser source He – Ne stabilized in frequency, having a resolution equal to 0,01 m.
The longitudinal comparator is based on the cinematic method, according to Abbe principle: the line scale must be aligned on the same longitudinal axis with laser beam on the measurement direction, in the scope to eliminate the first order errors. Because the line scale was short, it was supported with the whole surface on the special carriage of the comparator.
Measurement method:
Method of measurement consists in the displacement of the reference line measure along the measurement direction. Speed of displacement is about maxim 10 mm/min. Marks’ viewing of line measure is made using an optical microscope. The line is put on the mobile arm of the interferometer and the distance between marks is directly measured in length units.
The appearance of a fringe in the field of view of interferometer is determined by a deviation of optical path from measurement arm of interferometer equal with /2, where is the wavelength in the propagating medium at medium’s temperature, pressure and humidity of the laser He-Ne radiation, used as monochromatic source of light.
During one measurement between the beginning and the end of the measurement, difference between the thermometer readings was equal to tg = 0,01 0C. The difference between the temperature of the room and the reference temperature was 0,4 0C…0,8 0C during all measurements.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 80
EUROMET.L-K7-2006_Final-Appendix
LNMC Latvia
Action principle of horizontal comparator IZA-7 is based on optical viewfinder method.
Comparator IZA-7 design maintains adhering to principle of longitudinal comparison.
Measured item is possible to set up on the bedplate so that its axis could be the longitudinal extension of scale line axis. Scale line axis alignment with line of moving bedplate.
Measurement of length (line) is performed by method of comparison measured item line with comparator line scale using 2 microscopes with steadily span and parallel optical axis.
One microscope is used for aiming line, the other for reading on comparator scale.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 81
EUROMET.L-K7-2006_Final-Appendix
METAS Switzerland
Measuring system:
The measurements were performed on a 2D photomask measuring system with a measurement range of 400 mm x 300 mm. The system has an xy-stage with unique vacuum air-bearings featuring very small errors of motion. Two speed controlled servo motors move the table by means of fine strings. Once positioned, the stage is clamped to the granite base table by a vacuum brake. A differential two axis plane mirror interferometer (HP) measures the position of the stage. The moving mirrors are attached to a Zerodur base plate and the reference mirrors are fixed to the microscope objectives. Air pressure, temperature, humidity and CO2 content are on line accessed to determine the refraction index of the air by the Edlen formula. A microscope with a CCD camera and an episcopic illumination is used to localise structure positions. The microscope has a motorised turret and focus. For automatic focussing the image contrast is maximised.
The machine is located in a temperature stabilised clean room cabin of class 100 (US) within the clean room section here at METAS.
Scale mounting:
The line scales were aligned parallel to the x- or y-axis of the 2D photomask measuring system. They were supported at the Airy points, with a distance of 66.4 mm, by 4 spheres, two on a pivot. The scales were aligned using piezoelectric actuators. Vertically to better than 10 µm/100 mm. Additionally, as a 2D measuring system is used, the x-axis of the object coordinate system is placed trough the alignment marks by a numerical coordinate transformation to better than 4 µm/100 mm.
Line evaluation:
Each vertical line in the video image is analysed within the region of interest (ROI). The centre of a line profile is the average of the left and the right edge. The edge locations are determined
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 82
EUROMET.L-K7-2006_Final-Appendix
with a moment based edge operator. A line is fitted through all individual profile centres using only points within 2. The intersection of the fitted line with the reference line is used as the scale line position. The reference line was determined using the right and left pair of the horizontal alignment marks. Each line was evaluated within a length of 50 µm (height of the ROI).
Table: Image size and evaluation range.
Magnification 20x 50x
µm µm
Image size: x 244 98
y 182 73
Evaluation range: x 30 40
y 50 50
Typical line at 50x magnification with edge and line centre indication:
Measurement strategy:
The 2D photomask measuring system can operate fully automatically, therefore always a complete set of measurements was made lasting for about 2 hours. Such a set consisted of the following measurements:
Scale (mm) Lines Repetitions
0, 0.1 .. 1 11 10
0, 5 .. 100 21 10
0 .. 100 (for verification only) 2 10
Both line scales were measured in four orientations 0°, 90°, 180°and 270° with the 20x objective and additionally at 0°and 180° with the 50x objective. For the final result all these measurements were averaged. There was no systematic difference.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 83
EUROMET.L-K7-2006_Final-Appendix
Corrections:
For the final results all know corrections were applied. In particular, the thermal expansion was corrected using the given thermal expansion coefficients of 5E-07/K for quartz. The temperature deviations from 20°C were within ±0.1 K (average 20.06°C). Furthermore, as the measuring system is located 550 m above sea level, the average barometric pressure during the measurements was only around (945 .. 950) mbar. The results were reduced to the standard pressure of 1013 mbar with the given compressibility coefficient of -8.90E-07/bar for quartz. This correction was -5.9 nm for 100 mm.
Uncertainty contributions:
For the total uncertainty 30 contributions were considered. The largest contributions at 100 mm were for both scales, the repeatability, the air temperature and the scale support. For quartz scales also the material temperature is critical because glass has a low thermal conductivity and a small thermal capacity therefore it is difficult to measure its temperature. The accuracy of the thermistors itself is not the problem.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 84
EUROMET.L-K7-2006_Final-Appendix
MIKES Finland
Figure 1. MIKES' line scale interferometer
MIKES‘ line scale interferometer (fig 1) uses a dynamic method of measurement with a moving microscope for speed, simplicity, and considerations of space requirements. The graduation line distances are measured during continuous motion, which makes the system fast and the interferometer insensitive to minor turbulence in the interferometer beam path. Possible problems with speed fluctuations and time delay in observing the lines are avoided by using an electrically shuttered CCD camera as a line detector and synchronous data sampling.
The interferometer is constructed on a vibration isolated stone table to eliminate mechanical disturbances. The interferometer is situated in a laboratory room with air temperature 20±0.05 °C and humidity 46±2% RH. The microscope is fixed on one side of a carriage and the CCD camera on the top of the microscope, axis of which is adjusted perpendicularly to the scale plane.
The displacement of the microscope is followed by a Michelson interferometer utilising a calibrated 633 nm Zeeman-stabilised He-Ne laser. The light of laser is coupled via single mode fibre to the interferometer and adjusted parallel with the carriage movement by using a quadrant detector. Abbé error is eliminated with a large cube corner, making it possible to adjust the focus point of the microscope and the apex of the cube corner to the same point. This cube corner is constructed from three separate round mirrors adjusted to angles of 90 with each other. Ideal adjustment of the focus point and the apex of the cube corner nearly completely eliminates the Abbé error. The interference fringes are detected by two detectors with 90° phase difference and counted by a direction-sensitive quadrature counter.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 85
EUROMET.L-K7-2006_Final-Appendix
Integratedvideo sync.
Interferencesignal
Profiles of line images
N N
a1 b1 a2 b2
D
D
D
D
PP P
P
GRADUATION LINE #1 GRADUATION LINE #2
. . .
. . .a1
a1
b1
a2
a2
b2
b2
b1
1 2
Figure 2. Analysis of the measurement data.
In measurement run, the carriage moves in one direction while the programme continuously monitors the counter reading. When the carriage is approaching a line, the programme slows down the speed of the carriage and just before passing over the line it stores the current counter reading (Ni; figure 2), starting a synchronized sampling. In the sampling, one interference signal and an integrated video synchronisation signal, for determination of the field forming positions (ai, bi), are digitised, the graduation line image is stored, and the refractive index and temperature of the scale are calculated and stored. This set of samples is taken for each line after which a new run is started in the opposite direction. A single measurement of the decimetre lines of a 1 m scale takes approximately 15 minutes. The first approximation for the measured length is calculated as the distance between the positions where the graduation line fields are formed (Ni-Dij).
Average profiles of the graduation lines are formed by summing picture element intensities of each row of the CCD. Each image of the CCD camera consists of two fields charged in 1 ms and with time separation of 20 ms. Thereafter, the centre points of the graduation lines (Pa1, Pa2, Pb1, Pb2) are determined from the slopes of the line profile and a correction term needed to superimpose the centre points is applied. The refractive index of air is determined by Edlen’s formula updated by Bönsch et al. The line scale interferometer is capable of calibrating line separations from 10 µm to 1 m of good quality line standards, having line widths from 2 to 50 µm.
The measurements for comparison were done like for normal customer asking high accuracy calibration of good quality line standard. As specified in the protocol the width of the measuring section was adjusted to 502 µm and scale was supported from Airy points. The line spacings were measure in two different positions along the range of the stone rail. In both position the scale was measures in normal and reversed orientation. Each measurement consisted 5 runs. The averages of these 4 measurements are the results. Compressibility correction was calculated and applied.
Unfortunately the reference line was noticed to be non-ideal (fig 3). In separate test was studied what is the effect of this non-ideality. In test 10 first lines where measured normally and so that the defected area was excluded from the analysis. The difference was +12 nm with 6 nm STD. This correction was applied to the result. Also lines 0.1; 0.4; 85; 90 and 95 mm were clearly non-ideal (see receipt confirmation from MIKES). For this reason the value of the uncertainty component "Influence of line detection algorithm with line quality" was raised to 20 nm for these line. This approximately doubles their uncertainty.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 86
EUROMET.L-K7-2006_Final-Appendix
Figure 3. Non-ideality in the reference line (pointed by red arrow).
References:
- A. Lassila, E. Ikonen, and K. Riski, Interferometer for calibration of graduated line scales with a moving ccd camera as a line detector, Applied Optics 1994, 33, 3600-3603.
- A. Lassila, Updated performance and uncertainty budget of MIKES' line scale interferometer, In Proc. of euspen, Glasgow, Scotland, May 31 – June 2, 2004, p 258-259.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 87
EUROMET.L-K7-2006_Final-Appendix
MIRS Slovenia
Measurement setup
The line scale was measured on a Zeiss ULM 01-600 C 1D measurement machine by using laser interferometer HP 5528 A as a measurement system and Opto zoom microscope with CCD camera as a line detection system.
The scale is supported in Airy points by two metal half cylinders. Additional fixture is not necessary due to low movement speed (hand driven table).
The measurement setup is shown in Fig. 1
Fig. 1: Measurement setup
The centre of a line is detected by the video-positioning system consisting of a zoom microscope, CCD camera and a line detecting software, which was developed in the laboratory in the LabView environment.
Fig. 2: Line detection software – screen masks
The material temperature is measured with a Zeiss termistor contact sensor (resolution 1 mK, standard uncertainty of calibration 2 mK) on the measuring table, while the air temperature is measured with Agilent temperature sensor with resolution of 10 mK.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 88
EUROMET.L-K7-2006_Final-Appendix
Measurement procedure:
Each line position was measured from zero position (the system was positioned in zero point after each measurement). The whole procedure was repeated 10 times in a short time period. Arithmetic mean from those measurement represented measurement result, while the standard deviation represented the repeatability component of the measuring uncertainty. In order to evaluate reproducibility, the measurement was repeated in 5 different days with two different operators.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 89
EUROMET.L-K7-2006_Final-Appendix
NCM Bulgaria
The measuring system consists of a comparator, laser interferometer (with light source He-Ne laser 633 nm, HP5529A), moving carriage with mounting on it reflector and (with light source laser diode) for location of scale marks, temperature measuring system, barometer and hygrometer.
The scale was placed horizontally on two supports at Airy points.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 90
EUROMET.L-K7-2006_Final-Appendix
NML Ireland
Equipment Used:
Agilent 5519A Laser and SIP Horizontal Measuring Machine
Procedure:
The HP laser retro reflector is mounted to the end of the main bed of the SIP measuring machine, nearest the laser head. The HP interferometer is fixed to the longitudinal carriage of the SIP.
For line scales of nominal size greater than 10mm a transparent table is placed on the longitudinal carriage of the SIP measuring machine. This table is illuminated from below with a light source.
The glass scale under test is placed on the transparent table and aligned with the main axis (x axis) of the SIP using the adjustment screw on the transparent table.
The vertical microscope of the SIP is then aligned with the glass scale under test. An appropriate eyepiece and magnification is then selected to give optimum clarity to the glass scale under test, taking into account the line width and line quality of the line scale under test.
The central cross wire (the cross wires are imposed on the eyepiece of the SIP microscope) is aligned with the datum line of the glass scale. In this position the HP laser system is zeroed.
The carriage of the SIP is then moved until the second line of the glass scale is aligned with the cross wires. The distance moved is recorded from the display of the HP laser system.
This exercise is repeated at selected positions over the length of the glass scale. 5 repeat measurements are taken at each selected position along the scale and the average taken.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 91
EUROMET.L-K7-2006_Final-Appendix
NPL United Kingdom
The NPL linescale machine consist of a 400mm range, interferometrically monitored, air-bearing stage that moves the scale to be measured under a nominally fixed optical probe. The probe consists of the NPL NanoVision image processing system.
Scale motion is measured by two co-linear independent laser interferometers, one being an NPL differential Jamin type with plane mirror and second being an HP Michelson type with cube-corner reflectors. The measurement is carried out in a laboratory atmosphere and a weather station monitors the refractive index of the air. Following the Abbe principle, the common axis of the interferometers lies in the plane of the scale to be calibrated and is co-linear with the line containing successive measuring points on the scale. Separation of the cube-corner / plane mirror reflectors (one flexure-mounted) is maintained by a fused silica `wobble-pin'. The two beam splitter units are each flexure-mounted on the frame of the machine, separated from the measuring head by further silica wobble-pins.
A scale under examination is `Airy point' mounted with the chrome uppermost and with no compression force on the scale. The scale is illuminated with reflected light. A combination of the two measurements obtained greatly reduces the chances of common-cause error in displacement-measurement and the configuration provides an indication of accuracy and repeatability.
The 2006 measurements were taken with the NPL 400mm machine in a temporary laboratory ith poor environment conditions. The measurements were made without using an auto focusing routine.
The 2008 measurements were taken with the NPL 400mm machine in a dedicated laboratory in the New NPL building. The environment conditions here are much better. The measurements were made using an auto focusing routine and employing a second-generation version of NPL NanoVision video probe.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 92
EUROMET.L-K7-2006_Final-Appendix
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 93
EUROMET.L-K7-2006_Final-Appendix
NSCIM Ukraine
The measurement of the scale intervals was carried out by the absolute method using the interference setup of the primary standard of the unit of length DETU 01-03-98, which is a horizontal comparator with Michelson dynamic laser interferometer and photoelectric microscope for fixing the centre of the measured scale marks. The photoelectric microscope (PEM) and the corner reflector of the measuring arm of the laser interferometer are rigidly fixed on the movable table of the comparator. The scale under measurement is installed in a special positioner on the stationary table of the comparator.
To create the required measurement conditions the interference setup is placed into a heat chamber with automatically maintained air temperature of (200.05) оС.
The length of the scale intervals is determined in the wavelengths of the stabilized He-Ne laser radiation ( = 0.63 μm) by means of reversible counting of the fringes kept within two automatic settings of the PEM to the tops of the image of the measuring scale marks. The equation is as follows,
HtLnNL 160 ,
where N is the number of fringes (impulse number) kept within the measured scale interval; 0 is
the wavelength of the laser source in vacuum; n is the air refractive index; is the linear expansion factor of the scale material; t is the scale temperature deviation from 20 оС; HL is the nominal length of the scale interval.
The air refractive index is calculated by the Edlen formula using the measured values of pressure, temperature and humidity. The temperature of the gauges and the air is measured with regard to the base of the interference setup using the platinum resistance thermometer, copper constant thermocouples and direct current bridge. The pressure and humidity are measured using standard devices.
The real value of the scale interval length is determined as the average value of the measured scale interval lengths in the direct and reverse motion of the PEM.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 94
EUROMET.L-K7-2006_Final-Appendix
OMH Hungary
The measurements were done on the 3 m universal length measuring machine made by Zeiss, with the help of HP 5528 laser interferometer.
The environmental parameters were measured by independent calibrated measuring instruments. For the temperature measurements, 4 wires PT 100 thermoresistors were connected to Keithley type 196 multimeter. The readings were made automatically through HP-IB parallel interface. For the air temperature 2, for the material temperature 5 thermometers were used. The average of the 2 and the 5 values were used as air and material temperature.
The air pressure was measured by Wallace & Tiernan digital barometer Type Diptron 3, the data were taken also via HP-IB interface.
The air humidity was measured by a Digilog 60 capacitive digital humidity measuring instrument.
The correction factor was calculated by the EDLEN formula.
The data from the HP laserinterferometer were taken also via HP-IB interface.
The magnification of 100 X was used and the magnified image was seen on the SONY videomonitor through a videocamera. The image processing system generates digitally movable horizontal and vertical lines that serve to help to set the centre of the lines. The horizontal magnification was 1200 x.
The scale was placed on a steel base as the lines were standing vertically, 2/9L points were used.
The angle error of the carriage movement in vertical and horizontal plane (pitch and jaw) was measured by Agilent laserinterferometer in 0,1 sec resolution.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 95
EUROMET.L-K7-2006_Final-Appendix
PTB Germany
The PTB performed the measurements reported here with its Nanometer Comparator. Here only a short description will be provided. Detailed descriptions have been published already elsewhere [1-3]. A schematic of the Nanometer Comparator is shown in Fig. 1.
Figure 1: Schematic of the Nanometer Comparator
The Nanometer Comparator is equipped with a vacuum interferometer. An iodine-stabilized, frequency doubled Nd:YAG laser, which is operated according to the recommendations of the BIPM, serves as radiation source. The radiation is fed through polarisation maintaining fibres from the laser, which is located outside the measurement cabin, to the vacuum interferometer. The measurement slide is guided by air bearings and driven by a linear motor. Angular vacuum interferometers and piezo electric actuators integrated in the measurement slide are used in a closed loop to correct for the remaining angular deviations of the measurement slide. The z-axis actuators are also used to perform the z-axis fine motion required to determine the focus position of the sample. The sample is supported at its Airy points with a fixed and a movable roller. The fixed roller is supported by a height adjustable mount directly connected to the housing of the measurement reflector. The loose roller is put on a height stage mounted on the slide. A home-made optical microscope, which is shown schematically in figure 2, is used to detect the graduation lines of the scale. The housing was machined out of invar. The white light illumination is provided by a
Figure 2: Schematic of the microscope
cold light source and guided to the microscope by an incoherent fibre bundle. The illumination optics has been designed to achieve a Köhler illumination. The light is directed through the objective towards the sample by a pellicle. Here a 50 x Nikon measurement objective with a
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 96
EUROMET.L-K7-2006_Final-Appendix
numerical aperture of 0.55 and a working distance of 16 mm was used. By means of a fold mirror the image of the line can be observed directly using a CCD camera. This feature is used to align the sample horizontally. During the measurements the light of the image is detected by a photo diode after it passes through a fixed slit. In the object plane the slit has the dimensions of 4 µm × 50 µm. The intensity profile of the line and the related positions were acquired simultaneously by means of a hardware trigger provided by a function generator. In order to restrict the data acquisition to a region closely located around the line only, a gate signal for the trigger was derived from the digital scales of the measurement slide. The following algorithm was used to determine the position of a line. First the intensity profile of the line was normalized so that all values are between zero and one. Straight lines are then fitted to the intensity displacement dependence in the intensity range between 0.4 and 0.6. From this the positions, where the intensity equals 0.5, are calculated for right and left edge. Finally the average of these two positions is calculated and considered as the position of the line.
The measurements were corrected for the deviation of the air pressure and the temperature from the normal conditions (1013.25 hPa and 20°C) using the linear coefficients of the compressibility and thermal expansion as given in the technical protocol of the comparison. Furthermore the influence of the residual gas in vacuum chamber of the interferometer on the wavelength was corrected for using a simplified Edlen equation.
b.) Measurement conditions and measurement uncertainty
Due to the different spacings between the lines to be measured the measurements were divided in two independent parts. First the lines of the first millimeter and then the lines over the whole scale were measured. The scale was measured in the first and again in the third week of January 2007. The scale was aligned to better than 0.1 µm vertically and within 1 µm laterally the day before the measurements were performed. The whole setup then was left to reach its thermal equilibrium over night. Before the measurements were performed the focus position of the scale was determined and the z-position of the scale was readjusted accordingly. Each day 5 data sets in forward and backward direction were collected for the two independent parts. The measurement speed of the short part was 0.1 mm/sec and that of the long part was 0.2 mm/sec. Then the scale was remounted in the reverse orientation. To obtain the final result the average results of each day were averaged again. The results of the uncertainty estimation are summarised in the table on the following pages.
The pilot laboratory was informed in a separate short report about the influence of the partly contaminated graduation on the measurement results of the PTB (report sent on Feb 8th 2007). The severe contamination of the sample leads in general to a considerable increase of the measurement uncertainty. At several lines ( the position relative to the zero line are given in the table) an additional contribution was necessary to account for the experimentally observed scatter and the differences between the PTB mask comparator, which also has measured the scale, and the Nanometer Comparator. We like to mention here that neither the control measurements nor a comparison of these two instruments performed on the quartz scale used in the Nano 3 comparison performed parallel to the comparison showed any disagreements or variations, so that the observed larger variations of the results and differences between the instruments can be solely attributed to the contamination of the sample. In addition, a further comparison between the two instruments on a clean area of the scale also yielded a complete agreement within the observed experimental scatter, which was in the order of a few Nanometer.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 97
EUROMET.L-K7-2006_Final-Appendix
References
[1] J. Flügge, H. Dangschat, A. Spies, J. Tschirnich and H. Pieles: "Concept of a interferometric length comparator with measurement uncertainties in the nanometer scale", Proceedings of 1st Euspen Conference, Bremen, Mai 1999, p. 227-230
[2] J. Flügge, R. Köning and H. Bosse: "Status of the nanometer comparator at PTB", Proc. SPIE, Recent Developments in Traceable Dimensional Measurements, Jennifer E. Decker; Nicholas Brown; Eds., 2001, Vol. 4401, 275-283
[3] J. Flügge, R. Köning and H. Bosse: "Achievement of sub nanometer reproducibility in line scale measurements with the Nanometer Comparator", to be published in the SPIE proceedings of the Advanced Lithography conference, held in San Jose in February 2007
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 98
EUROMET.L-K7-2006_Final-Appendix
SMU Slovakia
The line scale was calibrated on the setup consisting from the 1-D length measuring machine Abbe Zeiss (measuring range up to 200 mm) and the laser interferometer HP 5529B.
The frequency (vacuum wavelength) of HP HeNe 633 nm laser has been calibrated against the HeNe/I2 633 nm primary standard of SMU.
The line scale was laid upon the 150 mm gauge block placed horizontally (height of 9 mm) supported by two semicylindrical supports in the mutual distance corresponding to Bessel points related to 150 mm long artifact of rectangular shape – the reason was possibility to use clamp attaching the Pt resistance thermometer directly to the gauge block onto which the measured glass artifact was resting.
The extrapolated axis of calibrated scale intersected the axis of cube corner and thus the Abbe principle was kept.
The carriage bearing both measured line scale and cube corner was moving under the optical microscope. Due to bending of the artifact over the whole measured range just the limited magnification could be used – objective 10x, eyepiece 12,5x
The position of the line was observed in the area determined by two parallel lines (at the begining and at the end of the scale) – as those as were not visible during the targeting on the measured lines, the height of observed area in the visual field was based on the subjective estimation.
The temperature of both the glass plate and air in the neighbourhood of the laser beam was measured by two Pt 100 thermometers (1 for scale, 1 for air), the air pressure was measured by the digital barometer DPi 141 and the relative humidity by the ALMEMO tester.
The scale was calibrated in two positions and the final result for each line is given by the arithmetic mean of two values corresponding to both positions of artifact (for every measured line, in each position the number of repeated measurements was equal)
The length values were corrected to 20 °C using the linear thermal expansion coefficient 5x10-7 K-1.
The pressure compressibility correction (negligible, cca -1 nm for 100 mm length) to 1013,25 hPa has been applied using the length compressibilty factor - 8,9x10-7 bar-1.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 99
EUROMET.L-K7-2006_Final-Appendix
ZMDM Serbia
The measurements were performed on a 1-D measuring machine (bench) with a total length of three meters (Zeiss type ULM 3000). Scale lines were detected using Zeiss optical microscope, mounted on a moving wagon of the measuring machine. A laser interferometer (HP 5526 A) was used to measure the displacement of the microscope.
Scale support: During measurements the line scale was supported on two roller bearings at Airy points without any clampings. Scale is aligned visually with the displacement axis of translation stage and measuring (laser) axis. Alignment is checked by comparing the focus and position of the end lines while displacing the translation stage from the line “0” up to the line “100”. This procedure is repeated several times.
Measuring bench: The translation stage of the measuring bench consists of two parts. The man part is sliding on the bench and is used to coarsely move the microscope into desired position. The second part of the translation stage (which is mounted on the main part) is used for fine-setting the microscope onto the line. Everything is operated manually.
Measurement procedure: The line scale was measured by coarsely move the microscope to each scale line (using the main part of the translation stage) and then fine-tune on the line with second part of the translation stage.
Since the instrument is manually operated, the measurements are time-consuming. In order to control induced drifts, the microscope was moved to the zero-point after every 10 lines measurements for checking.
Line scale was measured in two orientations with overall microscope magnification 50x. In one orientation two repeating measurements – forward and backward - were made as one group for each run. The total number of runs for each orientation were 12 except for total length 100 mm and first 1 mm (20). Then scale were reinstalled in reverse orientation and the alignement is checked as mentioned above. The final result is the average of all these measurements.
The positions of the lines were determined by direct visual positioning of vertical double cross-line of the measuring microscope at the center of each line in reflected white light with green filter. Left and right edge of each line were detected visually as the same (50%) gap light intensity level between microscope double cross-line and left and right edge of the line.
double cross-line
line on the measure
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 100
EUROMET.L-K7-2006_Final-Appendix
GROUP 2
A*Star – NMC Singapore
The comparison was done using a laser line width measurement system (as shown in Figure 1 & 2) at the National Metrology Centre, Singapore. The system consists of five major parts: a main base with two working stages driven by separate linear motors for sample holding and positioning, a laser interferometer for displacement measurement, a photoelectric microscope for line edge detection, an electronics unit for system control and a PC for data acquisition and recording. The He-Ne laser used in the interferometer is Lamb-dip stabilized and its wavelength is corrected for temperature, humidity and atmospheric pressure. The scale was placed so that the main inscriptions were uppermost and supported at the Airy points by stainless steel rollers. A lightly pinch was applied at the Airy points to avoid the scale shifting due to the movement of the working stage. The scale was viewed using transmitted light. The length of interval was measured along the longitudinal axis and at approximately mid-position of the scale. Measurements were taken in two orientations. At each orientation, the scale was properly aligned and two repeated measurements were made. For each measurement, it covered two independent measuring ranges: one from 0 to 1 mm and the other from 0 to 100 mm, both with a step of 0.1 mm. For the range from 0 to 1 mm, a closed reading was taken at the end of the measurement. For 0 to 100 mm, the measurement was divided into three sections: 0 to 50 mm, 40 to 90 mm and 80 to 100 mm in view of excess data collection. Adjustment factors at the overlapping intervals (40 to 50 mm and 80 to 90 mm) were considered in the calculation. The average of the four measurements of the two orientations was reported as the final results. All results of measurement are referred to a temperature of 20 C.
Figure 1: The laser line width measurement system
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 101
EUROMET.L-K7-2006_Final-Appendix
Figure 2: Enlarged view of the line edge detection unit
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 102
EUROMET.L-K7-2006_Final-Appendix
CEM Spain
Make and type of instrument
Custom-built length comparator CEM-TEK 1200 equipped with, two beams on flat mirrors, laser tracker and pseudo-Abbe measurement principle.
Light sources / wavelengths used or traceability path:
Stabilized Laser Source HP 5517C, 0 = 632.991 365 64 nm, calibrated against He-Ne reference laser CEM-1 (national standard)
Method used to determine the refractive index of the air :
Measuring of ambient conditions and applying of Edlèn’s formulae plus laser tracker (relative refractometer) for continuous updating of n value.
The laser is reseted in the first line (reference line position “0”) using a reference standard (gauge block) calibrated by interferometry.
Optical system:
The detection of the line is made by automatic recognition by the optical system with the help of home made software.
Alignment:
With horizontal and vertical micrometers, by observing the two main lines of the scale (at the beginning and the end) and the central one, trying to maintain in focus. For those less-in-focus lines, the optical system posseses autofocus.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 103
EUROMET.L-K7-2006_Final-Appendix
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 104
EUROMET.L-K7-2006_Final-Appendix
CENAM México
An optical microscope brand Leitz, model Libra 200, with 200x magnification was used.
The measurement range of the instrument is 150 mm x 100 mm, and the minimal division of the microscope scale is 0,1 µm
The lines edges were detected visually in a monitor connected to a CCD camera.
The distances between line centers were obtained by Micro/Measure Microscope Software and the temperature variation from the standard temperature was ± 0,4 °C.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 105
EUROMET.L-K7-2006_Final-Appendix
EIM Greece
The measurements were performed with the use of a Leitz universal measuring microscope with
x100 total magnification. A Renishaw laser system was used for the measurements of the line
scale displacement while the centers of the lines were estimated by the analysis of the digital
images of the line scale at each measuring position. The air temperature, humidity and pressure
were recorded during the measurement.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 106
EUROMET.L-K7-2006_Final-Appendix
INMETRO Brazil
The scale was calibrated in a system composed by a microscope mounted in an optical CMM (Trioptic–SIP) whose table is able to move in one direction. The table movement is detected by an interferometric laser system. The scale was laid over a special table so that it could be leveled aligned to laser beam and focused appropriately. Due to limitations, the scale was not supported in Bessel points as recommended in the Protocol. The measurements were performed in the same points as that ones outlined in the Protocol taking as reference the “0” point. The temperature during the measurements was (200,6) ºC, with a variation within 0,1 ºC.
2 – System configuration
Machine table
Optics and Laser Head
Leveling table
Microscope
CCD Camera
Glass scale
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 107
EUROMET.L-K7-2006_Final-Appendix
INRIM Italy
The measuring apparatus is based on a Moore Measuring Machine equipped with a laser interferometer and an optical probe. This latter consists of a optical microscope with a CCD-camera.
Tab. 1. Instrument identification.
Instrument Manufacturer Model Ser. No.
Universal Measuring Machine Moore n. 3 M245
Laser interferometer HP 5518 3626A03700
Microscope Nikon OPTIPHOT 100S 628562
Objective Leitz /0 Plan 125X N.A. 0,80
CCD camera Basler FM 60955
Measurement setup
The optical probe is used for the calibration of line standards. A rectangular window is created via software to simulate a probe; the “contact” is obtained from the image processing tool described below.
The window width corresponds to the ball tip diameter of a “contact” probe, whereas the window height determines the number of pixel rows activated (integration amplitude). By displacing the artefact (relative displacement between artefact and CCD camera), the window “penetrates” in the measurement area and defines the artefact edge position by measuring its distance from the window side (left or right, see Fig.1).
The optical probe is calibrated with reference to the laser interferometer. The pixel-size (about 0.08 µm/pixel, with a magnification of 125X) of the image window is obtained from the displacements measured with the laser interferometer when the window penetrates (of about 6 µm) the edge line from both left- and right-sides. In this way, the measurement runs are driven mostly as for the mechanical probe, except for the probe calibration which is based on the interferometer itself.
From the “contact” readings and the laser interferometer we determine the edge x-positions at the left- and right-sides of the line with an image window of about 32 µm width and 64 µm height. Then, the position of the centre of the line is obtained from the middle of the left- and right-edge positions.
Traceability is given by the wavelength of the laser interferometer.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 108
EUROMET.L-K7-2006_Final-Appendix
Fig. 1: The figure describes the way the position of the line was deduced from the image (a line of about
5 µm width is shown in the image)
Measurement procedure
The equipment configuration from bottom up was: Moore carriage, vertical stage (height adapter up to Abbe condition in vertical), tilt and rotary stages, base support designed for the Airy support points of the linescale, linescale.
The applied procedure is the following:
The linescale is placed on the base support and is aligned (visually) with the measuring axis. Then, the alignment is improved by checking the focus and position of the line end on the video frame while displacing the artefact from the line „0“ up to the line „100“ and back (to be repeated several times).
With the automatic control of the Moore machine, no manual handling of the linescale is required to reset the equipment between each set of measurements.
The positions of the lines are then measured at the central section of the lines with a CCD image window length of about 80 µm.
The adopted measurement strategy is: 1. Every 5 mm line over 100 mm (31 runs); 2. Every 0.1mm line over first 1 mm (16 runs);
About half of the measurement runs have been performed with the linescale in the reversed orientation.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 109
EUROMET.L-K7-2006_Final-Appendix
For each run the deviation from the nominal length is obtained from the average of forward and backward measurements of the line positions.
Ambient
Measurements have been carried out in a laboratory with air temperature control (20 ± 0.1°C). The table below gives the range of the ambient parameters during the measurement runs.
Ambient air temperature 19.96 – 20.08 °C
Pressure 99.7 – 100.8 kPa
Humidity ~ 50%
Air refractive index is calculated using the revised Edlen’s formula; the ambient parameters are measured by a precision thermometer, a Rosemount barometer and a Mitchell igrometer. The CO2 content is assumed to be 400 ppm.
B) Tabular description of the measurement methods and instruments
Line detection
Parameters Parameters used for the measurement
Microscope type: Nikon Optiphot 100S / Objective Leitz /0 Plan 125X N.A. 0,80
Light source Halogen lamp - Illuminator Intralux 4000
Wavelength(s) White light
window length 64 m
window width 32 m
Detection mode Image processing of the 2D CCD image video frame
Detection principle Left- and right-side line edge detection to determine the centre line position
Sampling frequency (image/interferometer)
15 frames/second; synchroneous reading of signal and interferometer
Edge detection criterion 50 % level dark-white light intensity
Edge detection short term repeatability (1s)
4 nm
Interferometric measurement
Parameters Parameters normally used for the measurement equipment
Achievable measurement uncertainty for measurands
Interferometer light source / wavelength
Stabilized He-Ne laser / 633 nm 7·10-9 · L
Resolution of displac. Interferometer
10 nm 2,9 nm
Interferometer medium Ambient air
Refractive index: calculated from ambient air parameters 2,3·10-8 · L
=> refractometer:
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 110
EUROMET.L-K7-2006_Final-Appendix
=> Edlen´s formula: revised Edlen’s Formula
Air temperature 20±0,1°C, precision thermometer
Air pressure ~ 100 kPa, Rosemount barometer
Air humidity ~ 50%, Mitchell igrometer
CO2-content Assumed to be 400 ppm
Other measurement conditions
Parameters Parameters normally used for the measurement equipment
scale orientations 0° and 180°
kind of support Airy support base
clean room class Not classified – laboratory with filtered air
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 111
EUROMET.L-K7-2006_Final-Appendix
NIM China
The line scale standards of No:2 was measured in Gauge Block and Line Scale Laboratory, Length Division, NIM. The comparator used for comparison measurement is made by NIM, which consists of He-Ne Laser with the wavelength 633 nm, main Interferometer, optical-electronic microscope, moving table, driving system, base, temperature measuring system, electronic part, computer and software. The interferometer is Mechlson type with resolution of 80 nm. The optical-electronic microscope uses dual slit to detect line position. Because the scale is transparent, the transmission light was used. Measurement starts from first line and continues till last line.
The principle diagram of Line Scale interference compactor
It took two weeks to finish total measurement. The kind of support is line contact. Four supports were used, whose positions were on the Bessel points marked on the scale. The scale was in series with the main beam path of interferometer with Abbe-error free. Scales were measured in two orientations. In one orientation two repeating measurements were made as one group. Then scale was reinstalled in reverse orientation and the instrument was readjusted such as focus and alignment and so on to eliminate possible system error. 48 groups of measurements were made, which means the Number of measurement of each scale is 96. The end result is the average of total measurement.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 112
EUROMET.L-K7-2006_Final-Appendix
Measuring environment condition is as following:
Measuring place: normal laboratory room
Room temperature: 20C±0.1C
Scale temperature: 20C±0.1C
Humidity: normal
Air pressure: 1015.5Pa±3.0 Pa
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 113
EUROMET.L-K7-2006_Final-Appendix
NIST USA
The 100 mm scale was measured with the N1ST Line Scale lnterferometer (LSI). The LSI consists of a scanning electro-optical line detector, a high precision one-axis motion system, and a high accuracy heterodyne interferometer for determining the displacement of the test artifact beneath the line detector. The wavelength of a stabilized helium-neon laser corrected for temperature, humidity, atmospheric pressure and CO2. is used as the length standard. The instrument is housed in an environmental chamber in which all environmental properties are carefully monitored. The complete description of the design and operation of the NIST LSI is given in the Journal of Research of the National Institute of Standards and Technology Volume 104, Number 3, May-June 1999, .. The NIST Length Scale Interferometer. ..
Reflected light was used to obtain line images. The scale was measured in the horizontal position with scale face up and supported at the Airy points. The scale was laid on a flat glass surface, supported at left in the center at a single point with a 6 mm diameter, 0.1 mm thick paper pad and the scale was supported at the right at two points, at the edges of the scale with similar pads. This support was used to avoid slippage of the scale, due to inertia during measurement, and the three point support prevents a twisting force on the scale as well as allowing free expansion and contraction of the scale.
On the 100 mm scale there were two independent measurements. From zero to 1 mm with 0.1 mm steps and from zero to 100 mm with 5 mm steps.
Measurements were made from line center to line center using a graduation line segment of 50 m long, in the center of the two horizontal alignment lines as indicated on page 11 fig.3 of the Technical Protocol document. Data was recorded by averaging 400 interferometer readings when the scale was stopped and servo-locked at each measured graduation line.
The environmental chamber and scale temperatures were held within ±0.005 °C of 20 °C during the measurements. The air temperature was measured close to the path of the interferometer laser beam and scale temperature was measured at three locations along the scale and the mean temperature was used for scale length corrections. The lengths are reported at a temperature of 20 °C (68 °F). A coefficient of linear thermal expansion of 5.1 x 10-7 / °C for the quartz scale were used in normalizing the lengths to 20 °C. During measurements the average atmospheric pressure was 99825 Pa and the average relative humidity was 50 %.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 114
EUROMET.L-K7-2006_Final-Appendix
Tabular description of the measurement methods and instruments
Line detection
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 115
EUROMET.L-K7-2006_Final-Appendix
Displacement measurement
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 116
EUROMET.L-K7-2006_Final-Appendix
Other measurement conditions
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 117
EUROMET.L-K7-2006_Final-Appendix
NMi-VSL Netherlands
Measurement set-up.
The line scale has been measured with the set-up shown in figure 1. The set-up consists of a SIP 400 measuring machine of which only the 400 mm travel of one of the axis is used for line scale calibrations. The alignment procedure and actual measurement is performed by manually adjusting the positions.
Figure 1 Overview of the line scale calibration set-up. The line scale is positioned on a specially designed translation stage (1b) of a SIP 400 length measuring machine (1a). The structures on the scale are visualized using a video camera (2a) and a video monitor (2b). After data processing of the video signal (2c) several control signals are displayed (2d) and used for alignment of the scale with respect to the camera. The position of the scale relative to the camera is measured using a laser interferometer (3a,3b) and stored on a computer (3c). Ambient conditions are constantly monitored and up to 10 sensors can be using to record (4a,4b) the temperature of critical components.
The original translation platform of the SIP 400 is replaced by a temperature controlled stage. The line scale and the retro mirror of the measurement arm are positioned onto an invar plate that is attached to the temperature controlled stage. The L-K7 line scale was supported on its Airy points.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 118
EUROMET.L-K7-2006_Final-Appendix
Figure 2 The user can select a region of interest in the field of view as indicated by the bright patch. Only information from the selected region is used for the alignment of the scale marker with respect to the camera system.
The markers on the line scale are imaged using a CCD camera and a video monitor. A 1000 x magnification is used to image the L-K7 scale structures. In order to obtain a picture of sufficient quality additional illumination is required that is provided by optical fiber along the camera mount. To minimize heating of the scale, the amount of light is controlled to a minimum.
The SIP table with the line scale is translated manually while the movement of the table is measured by a laser interferometer.
The alignment of the scale markers to the video camera is performed by processing the camera signal with a frame grabber and a computer. In the field of view the user can select a region that will be used for the measurement, see fig. 2. The image in the selected region is processed by averaging all image lines, normalizing the result and displaying this averaged information and its mirror image on a computer screen. Accurate alignment on a line scale marker is performed by maximizing the overlap of the averaged information and its mirror image, see figure 3.
Figure 3 User interface to align the line scale markers to the camera. The processed data from the selected region of interest (red line) is displayed along with its mirror image (yellow line).Depending on the measurement mode the alignment can be at the zero crossing or at a maximum or minimum.
The alignment can be optimized by maximizing the overlap of the two signals. When the operator has adjusted the position of the scale to optimized alignment the data from the laser interferometer is transferred to a second computer.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 119
EUROMET.L-K7-2006_Final-Appendix
The temperature of the air, line scale and SIP are continuously monitored during the calibration with a separate temperature measurement set-up. This set-up allows sequential recording of up to 10 temperature sensors. The temperature of the scale has been measured by positioning two thermistors on the scale surface above the Airy points. The thermistors were held by gravity without additional clamping. In order to keep the scale as clean as possible no thermally conductive paste was used to improve the thermal contact.
The alignment of the scale to the SIP is performed using the horizontal line structures at the left and right of the L-K7 scale. The alignment of the scale to the camera is optimized by matching the direction of the horizontal and vertical scale structures to software generated horizontal and vertical reference lines on the computer monitor. Final adjustments are optimized by rotating the line scale and/or camera and minimizing the vertical shift of the left and right horizontal lines while the scale is translated along its length.
Measurement procedure
The scale is measured both in the forward direction (0 mm to 100 mm) and in the backward direction. The center line position of the reference line (0 mm) is measured before each of the center line position of the line to be measured. During the acquisition the temperature of the air and of the scale, the air pressure and humidity are constantly monitored. These values are periodically updated when necessary (as judged by the operator) in the acquisition program for the laser interferometer. For this comparison the entire measurement has been repeated three times.
Analysis:
The analysis of the measurement data from the laser interferometer consists of averaging the values from the forward and backward series. The repeatability for a single measurement (forward and backward) is based on calculating the variance with respect to the average value for each position. The uncertainty due to the repeatability has been calculated from the square root of the averaged variances for three measurements.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 120
EUROMET.L-K7-2006_Final-Appendix
NPLI India
The glass scale is fixed on sliding platform of universal measuring machine (UMM) Model – MUL -214 B (Make SIP- Switzerland). It is aligned along the 50µm wide parallel lines looking into locating microscope of UMM. A heterodyne laser interferometer (LI), Model - Universal calibrator 5529A, (Make Agilent-USA), is set on the sliding platform of UMM to obtain measurements of graduations of glass scale.
The edges of the desired nominal graduation are detected by looking into locating microscope. Repeatedly, laser Interferometer is initialized at zero mark on the glass scale, then the platform of UMM is moved to locate nominal graduation of the glass scale under the locating microscope of UMM. The corresponding readings are measured by the LI and evaluated to determine the deviation (dL) from each nominal length.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 121
EUROMET.L-K7-2006_Final-Appendix
NRC Canada
The scale was measured on the NRC 4-metre Line Scale Comparator. This instrument has a granite bed (L,W,H = 4500,500,800 mm) with top and side surfaces lapped flat & straight. The aluminum carriage rides the length of the bed on air bearing pucks and is constrained by opposing air pucks that press against both sides of the bed. An HP heterodyne laser interferometer measures the displacement of the carriage, and a motor drive can position the carriage to within 1 µm of a requested position from the host computer.
Line Scale Set-up & Viewing: A fixed CCD camera/microscope is mounted on the carriage looking vertically down to a focus point 100 mm above the bed. Line scales to be measured are supported at this elevation above the bed, and are fine adjusted to be level and in focus during set-up by manual elevating jack/tilting screws. The line scales are also moved sideways with respect to the focus point, to centre the linescale measurand axis to the focus-point trajectory made as the carriage moves along the bed. Special jigs with these adjustment motions hold a long scale comprised of a rigid beam at the Airy support points, whereas short scales on thin substrates, such as microscope micrometer scales, are placed flat on an elevated glass-plate platform that allows either transmitted or reflective illumination.
For K7, the NRC Airy-point fixtures were not suited to directly support the 100 mm scale on its Airy points, and so it was simply laid flat on the invar platform support and fixed with plastic putty at its sides.
Displacement Interferometer: The carriage displacement interferometer uses an ‘open cube-corner’ design that allows the interferometer measuring point to coincide with the microscope viewing point, and thereby minimize Abbe error due to carriage motion error. On the front of the carriage are 3 small plane mirrors arranged to form the open cube-corner retroreflector, with its vertex made to coincide to with the microscope focus point. The HP interferometer optics are organized at one end of the granite bed. The displacement-measuring beam is directed along one side of the bed to the carriage, where it makes the succession of
reflections off the 3 mirrors and retro-reflects back down the bed along the opposite side: the coming & going beam paths are separated by 200 mm, so that large line scale objects up to 4 metres long can be laid between them. At the interferometer end of the bed is a small stationary cube-corner prism that reverses the path of the measuring laser beam, sending it back to the
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 122
EUROMET.L-K7-2006_Final-Appendix
carriage, traversing the 3 mirrors, and then back into the interferometer optics where it mixes with the reference laser beam to form the interferometer displacement signal. Thus the measuring beam makes a double-pass to measure the carriage displacement along the bed. The interferometer reference beam is also turned by a retro prism, but it also makes a double pass up-and-down the bed, this time up the middle to a plane mirror on a flexure anchored to the bed and contacting the end of the line scale to be measured: in this way the interferometer reference (zero) is coupled to the zero of the line scale with the smallest material ‘deadpath’ between them, and automatically compensates for any mechanical shift of the scale during measurement.
For the K7 scale mounted to the invar plate, the interferometer reference-arm retro-reflector was made to spring-contact the end of the invar plate.
Displacement Corrections: Sensors measure the air temperature, pressure and humidity to correct the laser wavelength in air, and also measure the line scale and the granite bed temperature to correct for material thermal dilatation. An autocollimator monitors the carriage motion error (pitch and yaw) so that residual Abbe corrections can be applied to the carriage displacement measurements.
Data Collection: The host computer executes a script of positions to visit along the scale, via the carriage motor drive. At each position the carriage is stopped and allowed to settle, then the CCD microscope image is recorded (actually, an average of 10 successive video images), along with the interferometer reading of the actual carriage position (actually, an average of 30 or more readings taken during the video capture) as well as all the sensor and autocollimator readings, plus a time/date stamp. This allows fast collection of data to minimize drift, and then processor-intensive offline analysis of recorded data runs can be done repeatedly and at leisure, allowing exploration of changes to analysis model parameters.
Line Center Algorithm: The CCD camera captures the microscope image at high magnification (typically 250 nm/pixel), and so the 50 µm ROI (region-of-interest) ‘slit height’ for K7 corresponded to 200 CCD lines in the image. The NRC target-center algorithm is based on the mean of left and right edge detection. The left and right edge of the intensity profile along each CCD line in the ROI was determined using a dynamic 50% threshold criterion, and the mid-position is calculated. A straight line, with outlier rejection, was fitted through the 200 mid-position points that spanned the height of the ROI (rejecting edge roughness or contamination artifacts better than a simple average), and the mid-height position of the fitted line was reported as the effective line centre.
Camera as Null Detector: As much as possible, line centres were always determined at same central location in the CCD image for every scale line, so that measured location of each line on the line scale depended only on the carriage displacement measurement by laser interferometer, and not the magnification scale factor of the CCD image. This was accomplished by moving the carriage in five small (3 µm) successive steps about the expected carriage position that put the desired scale line in the nominal centre of the CCD image: the carriage position was measured at each step (via the laser interferometer) and the image recorded. In offline analysis, the line image center was determined for each step, expressed in camera pixel coordinates, and a trend line fitted through line-image position (in pixels) vs. carriage position (in millimetres), to calculate the effective carriage position corresponding to that line imaged at the exact center of the camera field (pixel 320.00 in a 640 pixel field). This was done for every scale line in the same way, so that most residual systematic errors would be common-mode constants that cancel when all line positions are expressed relative to the zero-line. The residuals about the 5-point line fit were
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 123
EUROMET.L-K7-2006_Final-Appendix
typically less than 2 nm, suggesting that the fitted line centre location was consistent and repeatable to better than 2 nm.
Data Run Design: The script, or list of required measurement positions, is created offline in MS Excel. For the K7 scale, the list consisted the 30 prescribed lines plus the zero line. After the 100 mm line measurement, the carriage was driven 5 mm further, before reversing direction and measuring the 31 lines in reverse order. After returning to measure the zero line, the carriage was driven an additional 2 mm backwards, and then driven forwards to re-measure the zero line a final time in the data run, taken in the same direction as the first measurement, so as to reveal the closing error under hysteresis conditions. Thus, a K7 data run consisted of 31 + 1 +31 + 1 + 1 = 65 lines. This describes data runs in the normal orientation of the scale ( from 0 to the 100 mm line). A script was also designed for measuring the same lines when the scale was oriented in the reverse sense, starting a run at the 100 mm line and sequencing to the 0 line position, and then back to the 100 mm line (with the additional turn-around points as in the normal-orientation case.) Within a given data run, the forward-back repeatability of a given scale line location was the order of 20 nm, suggesting some drift problems, but differences the order of 100 nm were observed between normal and reversed orientation runs, as discussed below.
Data Runs for Reported Results: The results in the Table are the simple average of a normal-scale run, a reversed scale run and a 2nd normal scale run.
Observed Problems and Uncertainty: The NRC 4-metre Line Scale Comparator is still under development and has not yet achieved its expanded uncertainty performance goal of U = Q[20 nm, 0.2 ppm L], where Q[a,b] is the quadrature sum of a & b, and L is the distance between two scale lines (95% confidence interval). This uncertainty goal is the value expected when one applies a GUM model analysis to the customary influence factors for this type of comparator instrument, evaluated at the performance level of the modeled NRC instrument constituents.
However, during K7 data runs it was observed that after all corrections for environment, set-up and motion errors, there were still large differences between successive data runs and even larger between the normal and reverse data runs, calling for an increase in uncertainty to U = Q[80 nm, 0.4 ppm L] to account for the variations in the data runs. The source of the run-to-run and of the normal-to-reverse variations has not been identified yet, but may be due to failure to support the scale on the Airy points (causing unmodeled irregular bending distortions) and/or unmodeled thermal drift in the way the quartz scale is supported and fixed on the invar carrier, subject to illumination heating, and shifts/drifts in the coupling of the scale zero-line and the reference retro-reflector. NRC is investigating this aspect. Meanwhile, for the purpose of the K7 Report, NRC is not submitting a detailed GUM uncertainty model analysis while there exists such large rogue factors that dominate the performance.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 124
EUROMET.L-K7-2006_Final-Appendix
NIMT Thailand
Principle of measurement:
Graduated scales of the glass scale are measured using line scale interferometer. The light source is, wavelength 633 nm passes through the vacuum chamber and bellow.
Fig I. Diagram of vacuum laser interferometer system Resolution of the laser interferometer is approximately 0.8 nm (Referred to the manufacturer), which was calculated from (633 nm x 1/2 (optical part) x 1/400 (electronic part) 0.8 nm
Fig2. The system of the laser interferometer Graduated scales are detected by Edge sensor. Using triple slits system as show in Fig3.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 125
EUROMET.L-K7-2006_Final-Appendix
Fig3. Principle of Edge sensor Resolution of Edge sensor is approximately 0.64 nm (Referred to the manufacturer).
Fig4. Edge sensor detection Measurement setup:
Fig5. Support and setting
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 126
EUROMET.L-K7-2006_Final-Appendix
Temperature measurement:
Due to artifact is small, the temperature is measured at the standard scale instead.
Fig6.Temperature measurement Temperature measurement by: Precision Thermometry Bridge with probe Mfr: ASL, Model: F 300, Serial number: 00780303
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 127
EUROMET.L-K7-2006_Final-Appendix
VNIIM Rusia
An optical scheme of the comparator (Fig.1) consists of a laser polarization interferometer, a refractometer and a microscope. Parts of the laser interferometer and refractometer are fixed on a granite base. A carriage is moved over teflon supports at a distance of 1 m. As the carriage moves its rotation is not greater than 5 rad Two reflectors of the interferometer and the confocal microscope are mounted on the carriage. The microscope focus is located on the measurement axis of the inteferometer.
In the microscope there are a laser diode with a wave length of 540 nm and an objective with an aperture of 0.9 An illuminator of the microscope forms an illuminated strip with a width of 1m. Light is reflected and forms the image in a photodiode. The length of the slot corresponds to 50 m in the plane of a scale.
One of the two modes ( = 640 MHz) of the stabilized He/Ne laser is used in the interferometer, the second one is applied in the refractometer. A refractive index is measured in the process of pumping out a chamber of 1 m in length with air. Both the refractometer and interferometer have two photo diodes providing to shape the signals with a phase shift of 900.
The scale is located on two piezo-supports. Focusing is controlled by a signal from the microscope. The measurement zone is closed by a thermal screen. Alongside the scale a platinum thermometer (10 Ohm) is located. The difference between the temperature of the scale and that of the platinum thermometer is measured with a set of differential thermocouples. The temperature inside a room is kept at a level of 200.1 0 C.
Output signals of the photo-detectors of the laser interferometer, refractometer and microscope enter the computer that is equipped with an analogue-to-digital converter with a multiplexing unit at its input.
A phase of the interference signal is calculated as arctg of the ratio of the signals of two photo-detectors. Parameters of the input signals are corrected. The coordinates of the scale graduation line centers are calculated at the time of joint processing of the microscope and interferometer signals. A center of gravity is calculated for that part of the graduation line profile, which is situated between the levels of 25 and 40 % of a maximum level of the signal.
Measurements are done in a dynamic mode. When the microscope is moving over the graduation line the speed of the carriage is decreased (0.05 mm/s ).
Fig.1 Optical scheme of the comparator.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 128
EUROMET.L-K7-2006_Final-Appendix
Appendix 3: Reported uncertainty budgets
GROUP 1
BEV Austria
Input quantity xi Distrib. u(xi) unit i ci = dL / xi ui (dL)
pair N 15 Pa ∞ 2.7·10–9 Pa–1 L 0.04·10–6 L
tair N 0.15 °C ∞ 0.96·10–6 °C–1 L 0.14·10–6 L
cCO2 N 300 ppm ∞ 0.14·10–9 ppm–1L 0.04·10–6 L
nair N 5·10–8 ∞ 1 0.05·10–6 L
tS N 0.1 °C ∞ 0.5·10–6 °C–1 L 0.05·10–6 L
barometric compression L N 0.026·10–6 L ∞ 1 0.026·10–6 L
residual contributions Lres N 5 nm ∞ 1 5 nm
line pos. & Abbe & orient.
These errors can not easily be separated in our scheme. We use the standard deviation of 5 different calibrations (maximum of all lines) as explained on page 1 of this report.
Values are for the 1 mm and 100 mm range, respectively.
N 22 nm / 109 nm 4 1 22 nm / 109 nm
Since all contributions except the last one amount to only Q[5 nm, 0.17·10–6 L] it is tempting to assume a constant, length independent measurement uncertainty. Since the calibration procedure is somewhat different for short range (< 8 mm, mainly stage micrometers) and long range scales (> 8 mm) we use two constant uncertainties. The smaller one is valid up to 1 mm, the second one from 5 mm to 100 mm.
Combined standard uncertainty: uc(L) = 23 nm / 111 nm (@ 1 mm / @ 100 mm)
Effective degree of freedom: eff(dL) = 4 / 4
Expanded uncertainty: U95(dL) = 65 nm / 320 nm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 129
EUROMET.L-K7-2006_Final-Appendix
CMI Czech Republic
Input quantity xi Distrib u(xi) νi ci = dL / xi ui (dL) /
nm
Resolution of edge detection N 1.5 nm 1000 1 2 nm Repeatability of edge detection N 5 nm 30 1 5 nm
Interferometer resolution N 0.1 nm 1000 1 0 nm Interferometer nonlinearity N 1.0 nm 100 1 1 nm
Interferometer dead path influences N 0.8 nm 1000 1 1 nm Drift influence N 0.8 nm 100 1 1 nm
Influence of adjustment of measurement line + Nonlinearities of position sensing N 10 nm 20 0.577 6 nm
Microscope axis - ilumination alignment + defocusing N 5 nm 10 1 nm/um 5 nm
Influence of line edge detection algorithm, possible asymmetry of line profiles and shape N 10 nm 20 1 10 nm
Microscope magnification (or other position deviation sensing device) N 0.1 nm 20 138.0 nm/pix 7 nm
repeatability of scale preparation (bending by fixing, cleaning, …) N 12 nm 3 0.577 7 nm
δlwin Influence of adjustment of measurement window or slit length N 5 nm 5 0.707 4 nm
Errors due to Abbe offsets and pitch and yaw of translation stages at one position) N 1 nm 1000 1 1 nm
total constant part of u 17 nm
νeff 58 Possible contributions from interferometric displacement measurement technique: λo vacuum wavelength of light source used
for displacement measurement N 5.0E-11 1000 1 5.0E-11 nair Index of refraction of air2 N 1.0E-08 1000
tair Air temperature N 1.0E-02 20 9E-7 /K 9.0E-09 pair Air pressure N 5E-2 hPa 1000 2E-7 /hPa 1.2E-08
RHair Air humidity N 1.0E-02 1000 1E-8 /% 1.0E-10 cCO2 Air CO2 concentration N 100 ppm 1000 1.5E-10/ppm 1.5E-08
δlAi Errors due to Abbe offsets and pitch and yaw of translation stages N 1.7E-07 1000 1 1.7E-07
δlSi Errors of scale alignment N 2.5E-09 1000 1 2.5E-09 δlIi Cosine errors of interferometer alignment N 5.6E-09 1000 1 5.6E-09
systematic distortion of measurement frame with movement N 5.0E-09 1000 1 5.0E-09
Possible contributions from scale properties: αZ, Cr Linear coefficient of thermal
expansion of scale material N 0.0750 K 1000 1E-6/K 7.5E-08 κZ, Cr Linear coefficient of compressibility
of scale material N 750 Pa 1000 9.0E-12 /Pa 6.8E-09 δsupp Influence of support conditions N 2.3E-08 1000 1 2.3E-08
total relative part of u 1.92E-07νeff 1683
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 130
EUROMET.L-K7-2006_Final-Appendix
Combined constant part of standard uncertainty for one line position: uc=17 nm Combined constant part of standard uncertainty for one line distance from zero line, for three independent measurement (including removal and re-alignment) uc=14 nm
Combined standard uncertainty uc(L)=Q[14;192*L]nm, L in m
Effective degree of freedom: eff = 58 ; 1683
Expanded uncertainty: U95(L) = Q[28;377*L]nm, L in m
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 131
EUROMET.L-K7-2006_Final-Appendix
DZM Croatia
Input quantity xi Distrib. u(xi) unit i ci = dL / xi ui (dL) / nm, L in mm
Stage:
Error due to Abbe offset in z
(ap<0.5 mm) and pitch, δlAp R 10 100 1 10
Error due to Abbe offset in x, y
(ay<1 mm) and yaw, δlAy
R 15 100 1 15
Interferometer:
Interferometer wavelength, δλ R 0,03 100 L 0,03L
Air temperature, tair R 0,12 C 100 9,510-7 L/C 0,112L
Air pressure, pair R 13 Pa 100 2,710-7 L/Pa 0,035L
Relative Humidity, RHair R 0,06 100 8,510-7 L 0,050L
Edlen equation uncertainty, δnair N 210-8 100 L 0,020L
Interf. cosine error δlli R 0,8L 100 1 0,80L
Scale:
Scale temperature difference from 20 °C, ts
R 0,036 C 100 510-7L/K 0,018L
Errors of scale alignment, δlSi R 0,55L 27 1 0,55L
Movement of scale due to bending, δlai
R 0,2L 100 1 0,2L
Influence of line edge quality δEalg
R 10 49 1 10
Measurement:
Overall reproducibility, R N 55 27 1 55
Interferometer resolution, δlres R 6 100 1 6
Reproducibility of line detection, SE
N 35 27 1 35
Interferometer dead path influences, δlDP
R 30 100 1 30
Combined standard uncertainty: =Luc )( (75+0,5L) nm, L in mm
Effective degree of freedom: eff(dl) = 211, L=100 mm
Expanded uncertainty: U95(dl) =(150+1L) nm, L in mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 132
EUROMET.L-K7-2006_Final-Appendix
GUM Poland
Input quantity xi Distrib. u(xi) unit vi ci=dL/xi ui(dL) /nm
Line position sensing
Repeatability of edge detection sE N 130,0 nm 7 1 130,0
Reproducebility sR N 1 7 L L
Influence of adjustment of measurement line lpos N 30,0 nm 9 1 30,0
Interferometric displacement measurement
Vacuum wavelength of light source used for displacement measurement
o N 2,45·nm 100 L/ 0,00245L
Index of refraction of air nair N 1·10-8 100 L 0,010L
Air temperature tair N 51,6 mK 100 9,57·10-7·L/K 0,0494L
Air pressure pair N 6,3 Pa 100 2,68·10-9·L/Pa 0,0169L
Air humidity RHair N 0,462 % 100 8,67·10-9·L/RH% 0,004L
Air CO2 concentration cCO2 R 57,7 ppm 100 1,45·10-10·L/ppm 0,0084L
Interferometer resolution lRes R 2,9 nm 100 1 2,9
Interferometer nonlinearity lNL R 2,3 nm 100 1 2,3
Interferometer dead path influences lDP R 8,1 nm 100 1 8,1
Errors due to Abbe offsets and pitch and yaw of translation stages
lAi R 7,8 nm 100 1 7,8
Errors of scale alignment lSi R 1,2 nm 100 L 0,0012L
Cosine errors of interferometer alignment lIi R 18 nm 100 L 0,018L
Scale properties
Linear coefficient of thermal expansion of scale material R 2,89·10-8
1/K 100 L·0,14 K 0,004L
The difference of the scale temperature ts in oC during the measurement from the reference temperature of 20 oC
ts N 0,0173 K 100 5·10-7·L/K 0,0087L
Linear coefficient of compressibility of scale material
N 5,14·10-13
1/Pa 100 L·1170 Pa 0,0006L
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 133
EUROMET.L-K7-2006_Final-Appendix
Variations of air pressure during measurement ps N 6,29 Pa 100 8,9·10-12·L/Pa 0,0001L
Influence of support conditions supp R 0,6 nm 100 1 0,6
Combined standard uncertainty: uc(L) =Q [ 134 ; 1,00·L ] nm, L in mm
Effective degree of freedom: veff(dl) = 181 L = 100 mm
Expanded uncertainty: U95(dl) = Q [ 268 ; 2,00·L ] nm; L in mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 134
EUROMET.L-K7-2006_Final-Appendix
INM Romania
Input quantity xi Distrib. u(xi) unit i ci = dL / xi ui (dL) / nm
repeatability of the measurement
normal type A
0.05 m 19 1 50
thermal coefficient of expansion
rectangular
0.5·10-6 K-1
50 Lt
t = 0,5 0C
250L
temperature:
-calibration of thermometer
- reading error
normal type B
trapezoidal
0.025 0C
0.003 0C
8
L
25L
comparator certificate normal type B
0.05+0.1L, m 10 1 50+100L
wavelength correction due to the environmental parameters
normal type B
0.022L,m 50 1 22L
deviation of the flatness of the mirrors of retro-reflectors
rectangular
0.08 m 50 1 46
cosines angle rectangular
0.002L, m 50 1 2L
time of data acquisition rectangular
0.012 m 50 1 12
Line scale alignment rectangular
0.02 m 50 1 20
Rotation of the carriage rectangular
0.05 50 1 50
Combined standard uncertainty: =Luc )( [101,272L] nm;
Effective degree of freedom: eff(dl) = 100
Expanded uncertainty: U95(dl) = [202,544L] nm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 135
EUROMET.L-K7-2006_Final-Appendix
LNMC Latvia
Input quantity xi Distrib. U(xi) µm i ci = dL / xi ui (dL) / µm
Reference scale (0,03+2·10-7L) δls N 0,030 >100 1 0,009
Repeatability lR R 0,7 >100 1 0,7
Scale resolution δlX R 0,2 >100 1 0,12
Influence of focal length variation δEfoc
R 0,3 >100 1 0,17
Eccentricity coil of helical grid δlu R 0,2 >100 1 0,12
Difference in T between reference scale and comparator
T=±0,3ºC
α=(10,4·10-6) δt
R 0.3 >100 10,4·10-6 L 0,36
Deviation of the laboratory T from the reference scale
T=(20±2) ºC
±0.1·10-6 Δt
R 2 >100 0.1·10-6 L 0,012
Deviation of the laboratory T from the comparator
T=(20±2) ºC
±0.1·10-6 Δt
R 2 >100 0.1·10-6 L 0,012
Combined standard uncertainty: =(L)uc 0,823 µm
Effective degree of freedom: eff(dl) =
Expanded uncertainty: U95(dl) =1,647 µm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 136
EUROMET.L-K7-2006_Final-Appendix
METAS Switzerland
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 137
EUROMET.L-K7-2006_Final-Appendix
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 138
EUROMET.L-K7-2006_Final-Appendix
MIKES Finland
Input quantity xi Distrib. u(xi) unit i ci = dL / xi ui (dL) / nm
Repeatability(all components of type A n=20)
Std. 20.0 nm 19 0.23 4.6
Influence of line detection algorithm with line quality, Ealg
Std. 9 nm 4 1 9.0
Vacuum wavelength of laser, 0 Std. 4x10-9 100 1 L m 4.0 L
Index of refraction of air, effect of equation, nair
Std. 10x10-9 100 1 L m 10 L
Air temperature, tair Std. 0.015 K 4 960 L nm/K 14 L
Dew point of air, DP Std. 0.200 K 4 30 L nm/K 6.0 L
Air pressure, pair Std. 5.0 Pa 4 2.7 L nm/Pa 14 L
Air CO2 concentration, cco2 Std. 50 ppm 4 0.2 L nm/ppm 7.5 L
Linear coeff. of thermal expansion, Std. 500 nm/m/K 1 0.020 L Km 10 L
Scale temp. difference from 20 C, ts
Std. 0.010 K 2 500 L nm/K 5.0 L
Error due to Abbé offsets; pitch and yaw, lA
Std. 3.0 nm/0.1m 4 1 L 30 L
Effect of diffraction of laser beam, Std. 8 nm 2 1 L 8.0 L
Errors of scale alignment, ls Std. 1 nm 10 1 L 1.0 L
Effect of compressibility, Rect. 0.020 bar 4 900 L nm/bar 18 L
Effect of flatness dev. of the scale & microscope alignm., h
Std. 5 nm 3 1 5.0
Combined standard uncertainty: =(L)uc Q[11.3; 45 L] nm, L in meters
Effective degree of freedom: eff(L) = 12 => k=2.2
Expanded uncertainty: U95(L) = Q[25; 99 L] nm, L in meters
For lines 0.1; 0.4; 85; 90 and 95 mm the uncertainty due to line quality is higher than in the above budget based on non-idealities seen in the visual inspection (see receipt confirmation report from MIKES). The estimated expanded uncertainties for these lines are 54; 54; 55; 55 and 55 nm respectively.
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 139
EUROMET.L-K7-2006_Final-Appendix
MIRS Slovenia
Input quantity xi Distrib. u(xi) i ci = dL/xi ui (dL), nm, L in mm
wavelength drift, L R 2,410-9L 4 1 2,410-9L
Counting system R 5 nm + 2,510-8L 3 1 5 nm + 2,510-8LCorrection factor R 2,510-8L 3 1 2,510-8L
Wavelength correction for environmental influences, cor
R 2,810-7L 3 1 2,810-7L
uncertainty of determination of the scale mark center, s
R 26 nm 100 1 26
distance between the measurement and reference position in the video window, Lv
N 25 nm 100 1 25
LI indication in reference position, LLIref
N 25 nm 100 1 25
linear temperature expansion coefficient, m
R 1,1510-6C-1 100 1 C-1 L 1,1510-6L
temperature deviation, m N 0,1C 100 0,210-6 C-1 L 0,210-7L
cosine error, ecos N 0 100 1 0
random error caused by uncontrolled mechanical changes, eems
N 29 nm 100 1 29
measuring table inclination, ea R 30 nm 100 1 30
Combined standard uncertainty: =Luc )( (61+1,2L) nm, L in mm
Effective degree of freedom: eff(dl) = 520; L=100 mm
Expanded uncertainty: U95(dl) = (122+2,4L) nm, L in mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 140
EUROMET.L-K7-2006_Final-Appendix
NCM Bulgaria
Input quantity xi Distrib. u(xi)/ unit vi ci = ∂dL/∂xi ui (dL)/ nm
Vacuum wavelength 0 N 8.10-6 nm 100 L/ 13L
Index of refraction of air nair - 16.10-8 - L/n 160L
Scale temperature ts - 0,02 K - .L 10L
Linear coefficient of thermal expansion Z
- 4.10-8 1/K - (t-20)L 3L
Cosine error lli R 230 rad 5 0,4L nm/rad
92L
Error of scale alignment lSi R 405 rad 6 0,7L nm/rad
284L
Edge geometry influence Edef R 23 nm 4 1 23
Influence of line edge detection algorithm Ealg
R 17 nm 8 1 17
Error due to Abbe offset lAi R 40 nm 5 1 40
Influence of detection light wavelength lE
N 14 nm 10 1 14
Influence of focal length variation Efoc
R 35 nm 6 1 35
Influence of measurement in reversed orientation lRev
R 38 nm 5 1 38
Interferometer nonlinearity lNL
R 15 nm 5 1 15
Repeatability of edge detection sE
N 24 nm 20 1 24
Combined standard uncertainty: uc(L) = Q[78; 340L] nm, L in m
Effective degree of freedom: veff(di) = from 29 to 37
Expanded uncertainty: U95(dL) = Q[158; 690L] nm, L in m (k=2,03)
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 141
EUROMET.L-K7-2006_Final-Appendix
NML Ireland
Input quantity xi Distrib. u(xi) unit i ci = dL / xi ui (dL) / nm
Calibration uncertainty of Laser N 20 ∞ 1 20
Performance of Laser (1.7ppm) R 98.15 ∞ 1 98.15
Laser System Resolution R 5.77 ∞ 1 5.77
Temperature Corrections R 0.029 ∞ 750nm/˚C 216.51
Cosine Error R 3.87 ∞ 1 3.87
Line Scale Pitch & Yaw Error R 11.55 ∞ 1 11.55
Repeatability N 700nm 4 1 700
Combined standard uncertainty: =Luc )( 739.58 nm
Effective degree of freedom: eff(dl) = 5
Expanded uncertainty: U95(dl) = 1479.16 nm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 142
EUROMET.L-K7-2006_Final-Appendix
NPL United Kingdom
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 143
EUROMET.L-K7-2006_Final-Appendix
NSCIM Ukraine
Input quantity xi Distrib u(xi) i ii xdLc ixii ucu ,
nm
Inaccuracy of measuring the order of interference
N 14 nm 10 1 14
Wavelength reproduction error
R 1.8·10-9 100 L 0.002·10-6L
Inaccuracy of measuring the air parameters:
pressure
temperature
humidity
R
0.1 mm Hg
0.005 oC
0.1 mm Hg
100
100
100
0.357·10-6L
0.927·10-6L
0.056·10-6L
0.036·10-6L
0.005·10-6L
0.006·10-6L
Inaccuracy of measuring the scale temperature
R 0.005 oC 100 L
=5·10-7 1/ oC
0.003·10-6L
Inaccuracy of measuring the linear expansion factor of the scale
R 5·10-8 1/ oC 100 tL
t=0.1 oC
0.005·10-6L
Inaccuracy of the scale adjusting
R 5.8·10-4 rad
100 L 0.17·10-6L
Abbe error R 9.6·10-7 rad 100 1 10
Inaccuracy of determining the mark’s centre
R 10 nm 100 1 10
Combined standard uncertainty: 2)
61017.0(
2)20()( LnmL
cu
Effective degree of freedom: 29)( dLeff
Expanded uncertainty: 2)61017.0(2)20(295.0 LnmU
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 144
EUROMET.L-K7-2006_Final-Appendix
OMH Hungary
Input quantity xiDistributio
nu(xi) unit i
ui(dL)
[nm]
1 vacuum wavelength of HP 5528 laserinterferom Normal 6,0E-08*L 1 6,0E-08*L
2 Index of refraction of air nair Rectang. 1,0E-08*L 1 1,0E-08*L
3 Air temperature tair Rectang. 0,020 1E-06*L 2,0E-08*L
4 Air pressure Pair Rectang. 0,108 -3E-08*L -3,23E-08*L
5 Air humidity RHair (U=1,25%) Normal 0,169 1E-08*L 1,69E-10*L
6 Compensation factor of laserinterferometer Rectang. 0,058 1E-08*L 5,77E-08*L
7 Interferometer resolution LRes Rectang. 6 1 6
8 Interferometer dead path influences LDP Rectang. 11 1 11
9 Drift influence Rectang. 17 1 17
10 ABBE error in the xy plane Rectang. 23 1 23
11 ABBE error in the xz plane Rectang. 3 1 3
12 Cosine error of scale alignment in the xy plane Rectang. 3E-11 L 3,32E-11*L
13 Cosine error of scale alignment in the xz plane Rectang. 6E-11 L 5,89E-11*L
14 Cosine error of interferometer alignment Rectang. 3E-09 L 3,12E-09*L
15 Linear expansion coefficient of the scale material Rectang. 5,77E-07 -0,054*L 1,8E-08*L
16 Difference of the scale temperature from the refe Rectang. 0,11 -5E-07*L -5,53E-09*L
17 Length difference of scale position (standing or l Rectang. 5,00 1 5
18 Standard deviation max (*) Normal 94 19 1 94
ii xdLc /
*Note: In the line of 18: the maximum standard deviation is used (94 nm)
Combined standard uncertainty: =Luc )( Q[100; 9.4E-08 L] nm L=[nm]
Effective degree of freedom: eff(dl) = 25 (L=100 mm) k=2.11
Expanded uncertainty: U95(dl) = Q[211; 2E-07 L] nm L=[nm]
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 145
EUROMET.L-K7-2006_Final-Appendix
PTB Germany
for 100 mm const rel. to L
name and symbol xi (unit) Dist. xi ci = du(xi)/dxi ui(dl) /nm
ui(dl) /nm
ui(dl)*E9 /L
Translation stage:
Abbe offset y, z, Ay,Az = 1mm, 2mm, Yaw and Pitch of x-axis yaw (µrad) = 0.1 and Pitch (µrad) = 0.1
R
R
0.1
1 nm/urad
2 nm/urad
0.06
0.12
0.14
Interferometer:
Refractive Index ( due to rest gas, in hPa ) R 0.005 270 nm L/(m hPa) 0.078 0.78
Vacuum Pressure (hPa) R 0.001 270 nm L/(m hPa) 0.016 0.16
Interf. Deadpath LDP (20 mm), pressure stab. during measurement 0.001 mbar
R 20 270 nm LDP/(m hPa)
0.003 0.003
Interf. cosine error forw. beam
i (µrad) back. beam
R 15.2
30.54
-
-
0.0003
0.003
Beam diffraction, wbeam = 11 mm R 11 - 0.004 0.04
Vacuum wavelength 0 (/ = 1.4×10-11) R 1.4×10-11 L 0.001 0.01
Wavefront deformation N 1×10-9 L 0.1 1
Nonlinearities (nm) R 0.1 1 0.06 0.06
Microscope:
Defocus (observed dependence): 27 nm / 3�m, x=3×z2 nm/m2, zmax= 0.5 m
R 0.5 - 0.45 0.45
Microscope mode dependence R 15 1 8.7
Reference line R 15 1 8.7
Scale properties:
Scale alignment horiz. y = 0.5 m,
vert. z = 0.5 m
R 0.5
0.5
- 0.0004
0.0009
0.004
0.009
Scale alignment, position of support points, 1 mm interval
R 1 - 3.9
Unc. thermal exp. coef. u = 5×10-8/K, Tmax – 20 C < 0.06 K
R 5×10-8 0.06L×K 0.173 1.73
Temperature measurement, 10mK
= 5×10-7/K
R 10 0.5×L nm/(m K) 0.29 2.9
Pressure measurement upres= 0.6 hPa,
= -8.9×10-10/hPa
N 0.60 -0.89×L
nm/(m hPa)
0.06 0.6
Unc. Compressibility uҝ = 8.9×10-11 /hPa(p = 40 hPa)
R 8.9×10-11 40hPa L 0.21 2.1
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 146
EUROMET.L-K7-2006_Final-Appendix
for 100 mm const rel. to L
name and symbol xi (unit) Dist. xi ci = du(xi)/dxi ui(dl) /nm
ui(dl) /nm
ui(dl)*E9 /L
Measurement:
Reproducibility (4 nm, n=9) N 1.33 1 1.33 1.33
Contamination at lines (0.2,0.4,0.5,0.7) R 60 29
Total uc(1) 13 (31.8) nm
Total U95 30.8 (75.4)
General Expression U95 = 2.37× Q[13(32);4.1E-9×L]
Effective degrees of freedom: 8
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 147
EUROMET.L-K7-2006_Final-Appendix
SMU Slovakia
Combined standard uncertainty: =(L)uc 222 )094,0(46 L nm (L in mm)
Effective degree of freedom: eff(dl) ≈ 500
Expanded uncertainty: U95(dl) = 222 )19,0(92 L nm (L in mm)
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 148
EUROMET.L-K7-2006_Final-Appendix
ZMDM Serbia
Input quantity xi Distrib. u(xi) unit i ci = dL / xi ui (dL) / nm
Vaccum wavelength of laser, 0 R 410-9 100
1 L 410-9 L
Interferometer resolution, lRes R 3 nm 100 1 3
Index of refraction of air, nair N 110-8 100
2 L 110-8 L
Air temperature, tair R 0.1 C 8 9.510-7 L 9.510-8 L
Air pressure, pair R 30 Pa 8 2.710-9 L 8.110-8 L
Air humidity, RHair R 3% 5 110-8 L 310-8 L
Air CO2 concentration, cCO2 R 120 ppm 5 1.510-10 L 210-8 L
Linear coefficient of thermal expansion,
N 510-7 C-1 10 0.2 L 110-7 L
Scale temperature difference, ts
N 0.1 C 8 510-7 L 510-8 L
Interferometer alignment, lIi R 200 rad 15 xi2/2 L 410-8 L
Scale alignment, lSi R 200 rad 10 xi2/2 L 410-8 L
Errors due to Abbe offsets and pitch and yaw, lAi
R 20 nm 10 1 20
Microscope axis alignment, Ealig
R 30 nm 10 1 30
Repeatability of center line detection
N 80 nm 20 1 80
Drift influence (incl. deadpath) R 50 nm 20 1 50
Combined standard uncertainty: =(L)uc Q[101; 0.19L] nm, L in mm
Effective degree of freedom: eff(dl) = [42, 48L] L in m
Expanded uncertainty: U95(dl) = Q[202; 0.38L] nm, L in mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 149
EUROMET.L-K7-2006_Final-Appendix
GROUP 2
A*Star – NMC Singapore
Input quantity xi Distrib u(xi) unit i ci = dL / xi ui (dL) / nm
Vacuum wavelength, o R 310-8o 50 L/o 0.03L
Interferometer resolution, lRes R 24 nm 100 1 24
Cosine errors of interferometer alignment, lIi
R 230 rad 50 40010-6L/rad 0.1L
Air temperature, tair R 0.21 C 10 9.210-7L/C 0.19L
Air pressure, pair R 21 Pa 10 2.910-9L/Pa 0.061L
Air humidity, RHair R 2.3 %rh 10 8.910-9L/%rh 0.02L
Index of refraction of air, nair R 2.410-7 10 L 0.24L
Edlen equation, nEdl R 110-8 100 L 0.01L
Air CO2 concentration, cCO2 R 110-4 10 1.510-4L 0.015L
Errors due to horizontal Abbe offsets (h≤1mm), lAih
R 60 rad 50 h 60
Errors due to vertical Abbe offsets (v≤0.5mm), lAiv
R 60 rad 50 v 30
Errors of scale alignment, lSi R 50 rad 100 5010-6L/rad 0.003L
Interferometer dead path, lDP R 32 nm 50 1 32
Thermal expansion coefficient of scale, Z, Cr
R 310-8 C-1 50 0.5L/C-1 0.015L
Temperature difference of the scale from 20 °C, ts
R 0.2 C 10 510-7L/C 0.10L
Repeatability of edge detection, sE N 50 nm 13 1 50
Standard deviation of four repeat measurements, Lm
N 65.5nm 3 1 65.5
Combined standard uncertainty: =(dL)uc [(115)2 + (0.35L)2]1/2, L in mm
Effective degree of freedom: eff(dL) = 30 at L = 100 mm
Expanded uncertainty: U95(dL) = [(230)2 + (0.70L)2]1/2 (k=2)
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 150
EUROMET.L-K7-2006_Final-Appendix
CEM Spain
Input quantity xi Distrib. u(x
i) unit ν
i c
i = ∂dL / ∂x
i u
i (dL) / nm
Laser indication on reference line i N 22,2 nm 9 1 22,2
Laser indication on reference line 0 R 1,44 >100 1 1,44
Laser wavelength N 1,50e-09 >100 1,58e+03 L 2,37e-06 L
Index of refraction of air (initial reading of TRACKER)
N 3,52e-08 nm >100 1,00e+06 L 0,035 L
Thermal coefficient of line scale R 5,77e-08 ºC-
1 >100 1,00e+05 L ºC 5,77e-03 L
Temperature deviation of line scale from 20 ºC R 0,116 ºC >100 0,5 L ºC-1 0,058 L
TRACKER Accuracy (1) R 8,08e-08 L >100 1 0,081 L
Optics (2) R 3,7 nm >100 1 3,7 nm
Dead path R 8,1 nm >100 1 8,1 nm
Comparator misalignment (maximum for flat mirrors) (2)
R 2,89e-08 L >100 1 2,89e-02 L
Misalignment of the line scale R 1,02e-07 nm >100 1 1,02e-01 L
Optic set of line position N 15,3 nm 49 1 15,3 nm
(1) "Achieving Maximum Accuracy and Repeatability with the Agilent 5527A/B Laser Position Transducer System", Product Note.
(2) Optics and Laser Heads for Laser-Interferometer Positioning Systems, Product Overview, Agilent Technologies.
Combined standard uncertainty: uc (L) = Q [28.43; 0.15 L/mm] nm
Effective degrees of freedom: eff(dl) = 42
Expanded uncertainty: U95(dl) = Q [60; 0.3 L/mm] nm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 151
EUROMET.L-K7-2006_Final-Appendix
CENAM México
Input quantity, xi Distribution u (xi) vi
ii x
dLc
u2
i(dL) nm
Independent *L2
Repeatability of edge detection
N 120 11 (1 + scale*T)
14 293
Microscope magnification R 52 12 1 2 700 Microscope axis alignment R 4,4-7 100 L 0,19
Line edge detection R 29 100 1 833 Microscope calibration R 5,77 100 (1 + micro*T) 1 168
Linear coefficient of thermal expansion of
microscope scale
R
4,62-7
100
T*L
0,034
Difference of the
microscope temperature
R
0,29
100
micro*L
3,4
Linear coefficient of thermal expansion of scale
R
4-8
100
T*L
0,00026
Difference of the scale temperature
R
0,29
24
scale*T
0,026
Combined Standard Uncertainty: uc(L) = 2*8,318994 L
uc(L) in nm
L in mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 152
EUROMET.L-K7-2006_Final-Appendix
EIM Greece
Input quantity xi Distrib. u(xi) unit i ci = dL / xi ui (dL) / nm
Displacement Measurement R 1ppm 50 1 l
Artifact Temperature R 0.15 °C 50 a*l 0.075*l
Zero Line center estimate R 0.200 µm 50 1 200
Line center estimate R 0.200 µm 50 1 200
Scale alignment R 1.25 10-9*l 50 1 0.001 *l
L in mm
Combined standard uncertainty: =Luc )( 22283 L (nm), L in mm
Effective degree of freedom: eff(dl) = 123 for dl=100 mm
Expanded uncertainty: U95(dl) = 2*uc(L)
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 153
EUROMET.L-K7-2006_Final-Appendix
INMETRO Brazil
1. L from 0 mm to 55 mm Input Quantity xi Distrib. U(xi) unit i ci = dL / xi ui (dL) / nm
Repeatibility N 19 nm 9 1 19 nm
Resolution R 1 nm 1 0,577 nm
Variation in indication R 2 nm 1 1,155 nm
Laser alignment R 0,0001 nm 1 0,00005 nm
Dead length R 1 nm 1 0,788 nm
Errors of scale alignment R 0,02 nm 1 0,012 nm
Edge detection T 130 nm 1 53,072 nm
focal length variation R 50 nm 1 46,189 nm
Air temperature N 0,017 ºC 9,28E-07 (1/ºC) 0,434 nm
Air pressure N 10,9 (Pa) 2,68E-09 (1/Pa) 0,806 nm
Escale temperature N 0,010 ºC 5,00E-07 (1/ºC) 0,137 nm
Thermal expansion coeficient
N 0,000001 (1/ºC) 5,46E-01 (ºC) 15,015 nm
Partial vapor pressure N 16,374 (Pa) 3,71E-10 (1/Pa) 0,167 nm
--------------- --------- ------------- --------------- --------------
Combined standard uncertainty: uc(dl) = 63 nm ; 0mm ≤ L ≤ 55mm
Effective degree of freedom: eff(dl) = ∞
Expanded uncertainty: U95(dl) = 126 nm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 154
EUROMET.L-K7-2006_Final-Appendix
2. L above 55 mm mm to 100 mm Input quantity xi Distrib. u(xi) unit i ci = dL / xi ui (dL) / nm
Repeatibility N 10 nm 9 1 10 nm
Resolution R 1 nm 1 0,577 nm
Variation in indication R 20 nm 1 11,547 nm
Laser alignment R 0,0001 nm 1 0,00005 nm
Dead length R 1nm 1 0,788 nm
Errors of scale alignment R 0,02 nm 1 0,012 nm
edge detection T 200 nm 1 81,650 nm
focal length variation R 50 nm 1 28,868 nm
Air temperature N 0,017 ºC 9,28E-07 (1/ºC) 0,789 nm
Air pressure N 10,9 (Pa) 2,68E-09 (1/Pa) 1,465 nm
Escale temperature N 0,010 ºC 5,00E-07 (1/ºC) 0,250 nm
Thermal expansion coeficient
N 0,000001 (1/ºC) 5,46E-01 (ºC) 27,30 nm
Partial vapor pressure N 16,374 (Pa) 3,71E-10 (1/Pa) 0,303 nm
--------------- --------- ------------- --------------- --------------
Combined standard uncertainty: uc(dl) = 92 nm; 55mm < L ≤ 100mm
Effective degree of freedom: eff(dl) = ∞
Expanded uncertainty: U95(dl) = 184 nm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 155
EUROMET.L-K7-2006_Final-Appendix
INRIM Italy
name and symbol xi distrib. u(xi) unit i ci = dl/xi ui(dl) /nm
L in mm
edge detection repeatability between series
N 40 nm 30 1 40
interferometer digital resolution R 2,9 nm 100 1 2,9
vacuum wavelength N 5 fm 100 L 7·10-3 · L
air refractive index N 2,3·10-8 100 L 23·10-3 · L
interferometer non-linearity R 1 nm 50 1 1
interferometer deadpath R 2,4 nm 100 1 2,4
linear coefficient of thermal expansion () of scale
N 2·10-8 K-1 50 (t s - 20) · L 1·10-3 · L
difference of the scale temperature from the reference temperature during measurement
N 0,013 K 50 · L 6,5·10-3 · L
linear coefficient of compressibility (k) of the scale material
N 1·10-12 Pa-1 50 (101325 - p) L 1·10-3 · L
variations of air pressure during measurement
N 0,3 kPa 50 k · L 2,7·10-3 · L
Abbe offsets and pitch and yaw of translation stages
R 100 7
Imperfect alignment of scale (laser) with respect to meas. direction
R 100 50·10-3 · L
imperfect alignment in direction and height of the linescale
R 15 26·10-3 · L
Microscope axis alignment and straightness of translation stage
R 10 15 + 0,18 · L
Formula or expression of uncertainty shall be given in the same way as for MRA C:
Combined standard uncertainty: uc(dl) = 48 nm (L=100 mm)
Effective degree of freedom: eff(dl) = 50
Expanded uncertainty: U95(dl)/nm = Q[90, 0.4 L] ; L in mm
= 98 nm (L=100 mm)
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 156
EUROMET.L-K7-2006_Final-Appendix
NIM China
Input quantity x
i
Distrib.
u(xi) unit í
i c
i = ∂dL / ∂x
i u
i (dL) / nm
ë0 R 1.73×10-8ë0 50 dL/ë0 1.73×10-8dL
nair N 1×10-8 12.5 dL 1.00×10-8dL
tair R 0.0115 C 12.5 0.930×10-6dLC -1 1.07×10-8dL
RHair R 23.09Pa 12.5 0.371×10-9dLPa-1 0.86×10-8dL
Cco2 N 60×10-6 6500 1.45×10-4dL 0.87×10-8dL
Pair R 15Pa 12.5 2.683×10-9dLPa-1 4.03×10-8dL
SE N 0.48 39 ë0/8 38
Äts R 0.0058 C 12.5 5.0×10-7 dL C -1 0.29×10-8dL
áz,cr N 1.17×10-8 C -1 50 0.026 C ×dL 0.03×10-8dL
älDP R 0.8nm 2 1 1
älRes R 3nm ∞ 1 3
äEalig R 12nm 12.5 1 12
äEres R 0.289 ∞ ë0/8 23
älAi R 6nm 12.5 1 6
älNL R 12nm 12.5 1 12
älIi R 3.6×10-10dL 12.5 1 0.04×10-8dL
älSi R 2.4×10-10dL 12.5 1 0.03×10-8dL
D R 0.0116mm 50 1×10-7dLmm-1 0.12×10-8dL D-Diameter of diaphragm for incidence light
Combined standard uncertainty: uc = 282 )1080.4()48( dLnm =48.3¸nm≈49 nm
Effective degree of freedom: íeff
(dl) =95
Expanded uncertainty: U95
(dl) = 98 nm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 157
EUROMET.L-K7-2006_Final-Appendix
NIST USA
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 158
EUROMET.L-K7-2006_Final-Appendix
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 159
EUROMET.L-K7-2006_Final-Appendix
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 160
EUROMET.L-K7-2006_Final-Appendix
NMi-VSL Netherlands
Input quantity xi Distrib. u(xi) unit i ci = dL / xi ui (dL) / nm
Interferometer resolution R 0,4 nm inf 1 0,4 nm
Repeatability edge detection N 12 nm 93 1 12,0 nm
Wavelength
Laser R 3,1·10-9 inf L 3,1·10-9 ·L
Air temperature R 0,051 K inf 9,6·10-7 K-1·L 4,9·10-9 ·L
Air pressure R 22 Pa inf 2,7·10-9 Pa-1·L 5,8·10-8 ·L
Air humidity R 2,0 %RH inf 8,5·10-9 %RH-1·L 1,7·10-8 ·L
CO2 content R 58 ppm inf 1,5·10-10 ppm-1·L 8,4·10-9 ·L
Edlen equation R 1,0·10-8 inf L 1,0·10-8 ·L
Dead path (Ldp=10 mm)
Air temperature drift R 0.051 K inf 9,6·10-7 K-1·Ldp 0,5 nm
Air pressure drift R 5.8 Pa inf 2,7·10-9 Pa-1· Ldp 0,2 nm
Air humidity drift R 1.7 %RH inf 8,5·10-9 %RH-1· Ldp 0,1 nm
CO2 content drift R 58 ppm inf 1,5·10-10 ppm-1· Ldp 0,1 nm
Alignment
Scale to SIP 400 R 4,2·10-10 inf L 4,2·10-10 ·L
Laser to SIP 400 R 1,0·10-10 inf L 1,0·10-10 ·L
Interferometer optics R 0,1 nm inf 1 0,1 nm
Material temperature
Scale R 0,087 K inf 5,0·10-7 K-1·L 4,3·10-8 ·L
Expansion coefficient R 1,4·10-7 K-1 inf 0,015 K 2,2·10-9 ·L
Local heating R 0,194 K 150 5 nm·K-1 1,0 nm
Abbe error R 17,3 nm inf 1 17,3 nm
Combined standard uncertainty: =Luc )( 21 nm + 9,0·10-8 ·L
Effective degree of freedom: eff(dl) = 890
Expanded uncertainty: U95(dl) = 42 nm + 1,8·10-7 ·L
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 161
EUROMET.L-K7-2006_Final-Appendix
NPLI India
Source of uncertainty
xi ± u(xi) vi ci=dLi /xi ± ui(dLi)
nm
Edge detection
0 nm 144 nm 16.5 1 144
Alignment 0 nm 0.288 µrad/m Infinite L 288L
Thermal Expansion
5×10-7 K-1 2.8×10-8 K-1 Infinite L m × 0.5K 14L
Temperature variation
0 oC 0.28 K Infinite L m× 5×10-7 K-1
140L
Laser interferometer
100 mm 50 nm/m Infinite L m 50L
Repeatability 0 187 nm 20 1 187
veff. 36
t distribution 2 uc(dL ) 239
Combined standard uncertainty (k=1) uc(dL ) = ± [(236)2+ (324L)2]1/2
Combined standard uncertainty (k=1) uc(dL ) = ±239 nm at L = 100 mm
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 162
EUROMET.L-K7-2006_Final-Appendix
NRC Canada
No budget was reported. Only report in the form of tabular values for single measurement points was sent:
U = 80 nm (at 0,1 mm) to 89 nm (at 100 mm)
The following explanation was given:
“The NRC long-bed comparator was still under development at the time of measurement and an uncertainty model was not developed at that time. The cited uncertainty was based on repeatability & reproducibility data, plus allowances for some instrument problems we were suspecting at the time.”
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 163
EUROMET.L-K7-2006_Final-Appendix
NIMT Thailand
EUROMET.L-K7.2006 - Key Comparison: Calibration of line scales 164
EUROMET.L-K7-2006_Final-Appendix
VNIIM Russia
top related